Security Analysis Report of Kellogg (stock ticker K) Research Paper - 1

Security Analysis Report of Kellogg (stock ticker K) - Research Paper Example â€Å"The most common deciding factors used in the market to attempt to anticipate future stock pricing are recent price history and patterns, current price relative to past price levels or other stocks, beliefs about future price movement, and a fourth area that can be described generally as ‘reckless optimism† (Thomsett 1). 2. Discussion of background of the company. Kellogg is a leading producer and distributor of cereals, which was founded in the year 1906. From the year 1906 Kellogg had been expanding mainly by creating its own new brands of cereals, cornflakes, and other cereal related food products. The financial statement from 2006 to 2010 contains 5 years, which require to be inspected to recognize the stock activities in Kellogg. It is listed on the Dow Jones Stock Exchange and New York Stock Exchange. Kellogg’s had expanded over the years and has a good market share. Its stocks and share prices showed a very positive trend and investors were on a rush t o invest in Kellogg stocks. Due to economic recession of 2007 like other companies kellogs socks also took a stagnat approach. So this paper investigates whether an investigator should invest in Kellogg’s shares by buying, or he should be holding his decision or in fact the investor should be selling Kellogg’s shares in the stock market. ... A stock is considered as best or perfect only when all the dimensions are perfect and project a future growth for its investors. An expanding business has growing and healthy revenue. The stock movements are a projection of a business and past results are no guarantee yet the past stock records of a firm influence its present and future movements to a certain extent. A firm which has high sales does not mean it is earning high profits. High sales do not measure a stock’s performance, whereas the profits of a firm directly affect the stock movements. From the income statement of Kellogg, it is understood that the net sales from the year 2006 to 2008 show an increasing trend while the year 2009 and 2010 point out to a decreasing movement in net sales as compared to the previous year. From 2006 to 2008 there was an increase in net sales by around 18%. While in the year 2009 there was a drastic downfall by around 2%, similarly, the year 2010’s net sales also showed a decrea se by 1.41% A consolidated average from the year 2006 to the year 2010 with respect to the net sales of the firm shows that Kellogg has experienced a growth of 13.66%. Comparing the net earnings of Kellogg there is an increase of 24%. In such a case, Kellogg stocks have to be brought. Comparing the net earnings of the years 2009 and 2010, there is a decrease of 2.9%. The profits of a firm are understood from its payment of dividends. A firm with a strong balance sheet does not have to worry about its debt. The return on equity helps to measure how well a firm is converting its resources into profitable business endeavors for expansion purpose. If the five-year annual revenue growth is more

Ways in which Arthur Miller creates tension in the first act Essay Example for Free

Ways in which Arthur Miller creates tension in the first act Essay In this essay I am going to explore the ways in which the writer, Arthur Miller, creates tension in the first act of The Crucible. Some of the techniques he uses, and I am going to analyse are: pace, fear of witchcraft, disagreements, and the relationships between the characters. The play includes moments in which the pace is slow, this provides a contrast to the moments of climax, when the pace quickens. As there is a change in pace which the readers and viewers cant expect, the tension increases. For example, the scene in which a psalm is gently sung when, suddenly, Betty starts screaming. The play begins steadily and calmly, no tension is thought to be created. The upper bedroom in the home of Reverend Samuel Parris is slowly described, it gives the impression of being a peaceful place. There is a narrow window at the left. Through its leaded panes the morning sunlight streams. A candle still burns near the bed, which is at the right. Some characters introductions and speeches also make the action go slower. Some examples are the introduction of Reverend John Hale Mr Hale is nearing forty, a tight-skinned, eager-eyed intellectual. and the experienced, comforting speech of Rebecca Nurse when Mr Putnam asks her to see if she can wake up his daughter: I think shell wake in time. Pray calm yourselves. I have eleven children, and I am twenty-six times a grandma, and I have seen them all through their silly seasons ( ) These slow scenes give extra emphasis to the parts in which the speed of the scene increases by the things that happen or people say. Any important exclamation in any dialogue or threats to other characters can make the pace change in this way. Some examples are: Abigail Williams threatens Mercy and Betty to be quiet about what really happened the night on the forest.

Multilevel Thresholding According to Histogram

Multilevel Thresholding According to Histogram Make Multilevel Thresholding According to Histogram by Cooperative Algorithm based on AFSA and Fuzzy Logic Image segmentation is a technique which is usually applied in the first step of image analysis and pattern recognition and is an important component of them. This technique is taken into account as one of the most difficult and the most sensitive problems in image analyzing. In this paper, a cooperative algorithm is proposed based on AFSA and k-means. The proposed algorithm is used to make multilevel thresholding for image segmentation according to histogram. In the proposed algorithm, first, artificial fish (AF) perform optimization process in AFSA. After swarm convergence, obtained cluster centers by AFs are used as initial cluster centers of k-means algorithm. After forwarding AFSAs output to k-means, AFs are reinitialized and performs clustering again. The proposed algorithm is used for segmenting 2 well-known images and obtained results are compared with each other. Experimental results show that segmented images quality by the proposed algorithm is much better than four other t ested algorithms. Keywords: Multilevel Thresholding; Histogram; Cooperative Algorithm; k-means. Image segmentation is a technique which is usually applied in the first step of image analysis and pattern recognition and is an important component of them. This technique is taken into account as one of the most difficult and the most sensitive problems in image analyzing. In fact, quality of final result of image analysis depends highly on the quality of image segmentation result. In image segmentation process, an image is divided into different regions. Segmentation approaches of mono-color images are with respect to discontinuity and/or similarity of gray level amounts in one region. If the approach performs segmentation based on discontinuities, the image is segmented with respect to abrupt changes on gray level by means of recognizing dots, lines and edges [1].The purpose of image segmentation approaches is to classify and convert pixels into regions. Histogram thresholding is one of the techniques, which has been applied extensively in mono-color images segmentation [2]. Generally, images are composed of regions with various gray levels. Therefore, an images histogram can consist of some peaks that each of them is related to one region. To separate boundaries of two peaks from each other, a threshold value is considered between valleys of two adjacent peaks. Indeed, histogram thresholding is a famous technique which is looking for peaks and valleys in a histogram [3]. Various clustering algorithms such as k-means [4] and FCM [5] have been used for histogram thresholding so far. As a matter of fact, clustering approaches, because of simplicity and effectiveness, belong to the most famous techniques that could be used for natural image segmentation. Applying clustering algorithms in histogram thresholding are such that first colors histogram is built and after that, clustering is done according to color distribution among pixels. O ne of the clustering methods is to use such swarm intelligence algorithms as particle swarm optimization (PSO) [6], and artificial fish swarm algorithm (AFSA) [7]. PSO was presented by Kenedy and Eberhart in 1995 [8]. Different versions of this algorithm have been used many times in data clustering [9]. Artificial fish swarm algorithm (AFSA) was presented by Li Xiao Lei in 2002 [10]. This algorithm is a technique based on swarm behaviors that was inspired from social behaviors of fish swarm in nature. AFSA works based on population, random search and behaviorism. This algorithm has been applied on different problems including machine learning [11, 12, 13], PID controlling [14], image segmentation [16], data clustering [7, 16] and scheduling [17]. K-means or famous Lloyd algorithm is one of the famous data clustering algorithms [18]. This algorithm is of high convergence rate, but has some weaknesses such as sensitivity to initial values of cluster centers and convergence to local op tima. Researchers have tried to remove these weaknesses by hybridizing this algorithm with other algorithms such as swarm intelligence ones [6, 19] and to utilize their advantages. One of these algorithms is KPSO in which first, k-means is performed and after that outcome of k-means is delivered to PSO as a particle [20]. Hence, at the beginning of the algorithm, k-means reaches to a local optimum with its high convergence rate and after that PSO takes the responsibility of increasing the result accuracy and exiting form local optimum. In this paper, a cooperative algorithm is proposed based on AFSA and k-means. The proposed algorithm is used to make multilevel thresholding for image segmentation according to histogram. In the proposed algorithm, first, artificial fish (AF) perform optimization process in AFSA. After swarm convergence, obtained cluster centers by AFs are used as initial cluster centers of k-means algorithm. After forwarding AFSAs output to k- means, AFs are reinitialized and performs clustering again. In fact, in the proposed algorithm, AFSA is used for a global search and k-means is used for a local search. The proposed algorithm along with four other algorithms is used for image segmentation on two known images Lenna and Barbara. Efficiency comparison shows that the proposed algorithm has an appropriate and acceptable efficiency. The remainder of the paper is organized as follows: in sections 2 and 3, standard AFSA and k-means algorithm will be described respectively and in section 4, the proposed algorithm will be presented. Section 5 studies the experiments and analyzes their results and final section concludes the paper. In water world, fish can find areas that have more foods, which is done with individual or swarm search by fishes. According to this characteristic, artificial fish (AF) model is represented by prey, free-move, and swarm and follow behaviors. AFs search the problem space by those behaviors. The environment, which AF lives in, substantially is solution space and other AFs domain. Food consistence degree in water area is AFSA objective function. Finally, AFs reach to a point which its food consistence degree is maxima (global optimum). In artificial fish swarm algorithm, AF perceives external concepts with sense of sight. Current position of AF is shown by vector X=(x 1, x 2,à ¢Ã¢â€š ¬Ã‚ ¦, x n). The visual is equal to sight field of AF and Xv is a position in visual where the AF wants to go. Then if Xv has better food consistence than current position of AF, it goes one step toward X v which causes change in AF position from X to Xnext , but if the current position of AF is better than X v, it continues searching in its visual area. Food consistence in position X is fitness value of this position and is shown with f(X). The step is equal to maximum length of the movement. The distance between two AFs which are in Xi and Xj positions is shown by Dis ij =||X i-Xj|| (Euclidean distance). AF model consists of two parts of variables and functions. Variables include X (current AF position), step (maximum length step), visual (sight field), try-number (the maximum test interactions and tries) and crowd factor ÃŽÂ ´ (0 The standard k-means algorithm is summarized as follows: Initial position of K cluster centers is determined randomly. The following steps are repeated: a) for each data vector: data vector is allocated to a cluster that its Euclidean distance from its center is smaller than the other clusters centers. Distance from cluster center is calculated by Equation (1): (1) In Equation (1), Xp is data vector p, Zj is the center of cluster j and d is the number of dimensions of data vectors and cluster center vectors. b) After allocating all data to clusters, each of cluster centers is updated by Equation (2): (2) Where, nj is the number of data vectors that belong to cluster j and Cj is a subset of all data vectors which belong to cluster j. The resulted cluster center of Equation (2) is the average vector of data vectors comprising cluster. (a) and (b) steps are iterated until the stopping criterion is satisfied. In this section, the proposed algorithm is described. In the proposed algorithm, there exists a population of AFSAs AFs. This population of AFs is initialized randomly in problem space. Each AF consists of K cluster center positions in one dimensional image histogram space. Therefore, search space for AFSA for K cluster centers has K components. Fitness function which AFSA has to minimize is shown in Equation (3). (3) Clustering on histogram is done by Equation (3) based on color distribution between given images pixels. The image is divided into K clusters (Ci) according to color attribute by K-1 thresholds. In Equation (3), the distance between color Xj on image histogram and the center of a cluster which it belongs to ( Zi), is multiplied by the frequency of pixels (fj) which have color value Xj on given image. This value is computed for all color values with respect to the center of a cluster which they belong to. Each color becomes the member of a cluster in which their distance from that cluster center is less than other cluster centers. Finally, the obtained results of all clusters are summed with each other. Indeed, Equation (3) calculates sum of intra cluster distances for one dimensional gray scale images, which is one of the most well-known clustering criteria. For improving obtained results by AFSA, some modifications must do on its structure. The best found position by swarm members so far in AFSA is saved in bulletin and AF which has found it might go even toward worse positions with performing a free-move behavior. Therefore, AFs cannot utilize their best swarm experience for improving the convergence rate because they just save it in bulletin. On the other hand, performing free-move behavior is inevitable for maintaining diversity of the swarm. In this paper, to remove this problem, every AF except best AF can perform free-move behavior. In fact, during execution of the proposed algorithm, this behavior is not performed for the best AF of the swarm at all. Hence, the best found position by the swarm would be the position of the best AF of the swarm. As a result, other members of the swarm can move in the direction of the best found position by executing follow and swarm behaviors. The purpose of designing the proposed algorithm is to take advantages of both AFSA and k-means algorithms and remove their weaknesses. K-means is of high convergence rate, but its very sensitive to initializing the cluster centers and in the case of selecting inappropriate initial cluster centers, it could converge to a local optimum. AFSA can pass local optima to some extent but cannot guarantee reaching to global optima. However, AFSAs computational complexity for optimization process is much more than k-means. How the proposed algorithm functions remove weaknesses of these two algorithms and apply their advantages is as following: In the proposed algorithm, first, the AFs are initialized in AFSA. Each of AFSA contains K cluster centers (K-1 threshold) which are displaced in the problem space by performing AFSAs behaviors. AFSA continues to perform until the AFs converge. After convergence of AFSA, best AFs position including the best cluster centers which have found by AFs so far is considered as the input of k-means. Then, k-means algorithm starts working and while it is not converged, it continues working. Therefore, AFSA searches globally and as far as it can, it passes local optima. After convergence of AFSAs AFs, its output would have an appropriate initial cluster centers for k-means. Hence, after sending AFSAs outcome to k-means, this algorithm starts searching locally. Consequently, in the proposed algorithm, global search ability of AFSA has been used and after converging, a great part of optimization process will be given to k-means to utilize high capability of local search of this algorithm and its high convergence rate. Since initial cluster centers for k-means are obtained by AFSA and k-means is used for local search, k-means weakness of sensitivity to initial cluster centers is removed. But, AFSA capability may not be enough for preventing from being trapped in local optima. If this algorithm is trapped in local optima, it cannot present proper initial cluster values to k-means. Thereafter, according to low ability of k-means in passing local optima, the obtained result cannot be acceptable. To raise this problem, after convergence of AFSA, the output of this algorithm is sent to k-means. Simultaneously with starting of k-means, AFSAs AFs are initialized and start global search again. In fact, in one time of executing the proposed algorithm, AFSA has several times of chance to perform an acceptable global search. It should be noted that in the proposed algorithm, in each time of executing AFSA, AFs just search globally and converge after a short time and k-means undertakes the remaining of optimization process which is local search. Therefore, with respect to low computational complexity of k-means, huge amount of computations for local search is prevented. In the proposed algorithm, it has been tried to utilize this conserved computation load for giving new opportunities to AFSA in order to perform an acceptable global search in at least one of given opportunities to it. Hence, for each execution of global search by AFSA, k-means is also performed once. In the proposed algorithm, to determine the convergence of artificial fish swarm, the difference of obtained results in consecutive iterations of performing the algorithm is used. When particles converge, the obtained results difference in consecutive iterations decreases, so by considering a threshold for the difference between best AFs fitness values in iterations i and j, it can determine their convergence. In the proposed algorithm, because AFSA and k-means algorithms are performed multiple times , always, it has to save the best found cluster centers by algorithm so far. For this purpose, a blackboard is applied that each time k-means finishes after convergence of AFSA, the obtained result of that will be compared with saved result in blackboard. If obtained cluster centers are better than saved result in blackboard, saved value in blackboard is updated. K- means execution finishes when after two consecutive iterations of its execution, cluster centers wouldnt be displaced. Pseudo code of the proposed algorithm is represented in Figure (1). Experiments are done on two known gray scale images, Lenna and Barbara, of sizes 512*512 in Figure (2). In this paper, the well-known criterion of uniformity is used to compare images segmentation qualitatively [3] which is shown in Equation (4) (4) Where, c is the number of thresholds. Rj is the segmented region j. N is the total number of pixels in the given image, fi shows the gray level of pixel I,  µi is the mean gray level of pixels in jth region, finally, fmin and fmax are the minimum and maximum gray level of pixels in the given image, respectively. Usually, uà Ã‚ µ[0, 1] and larger amount for u declares that the thresholds are specified with better quality on the histogram. Proposed Algorithm: 1:for each AFi 2:initialize xi 3:Endfor 4:Blackboard = arg [min F(Xi)] 5:Repeat 6:for each AFi 7:Perform Swarm Behavior on Xi(t) and Compute Xi,swarm 8:Perform Follow Behavior on Xi(i) and Compute Xi,follow 9:if F(Xi,swarm) à ¢Ã¢â‚¬ °Ã‚ ¥ F(Xi,follow) 10:then Xi(t+1)= Xi,follow 11:Else 12:Xi(t+1)= Xi,swarm 13:Endif 14:Endfor 15:if swarm is converged 16:then Execute k-means on XBest-AF until stopping criterion of k-means is met 17:Endif 18:if F(Xk-means) à ¢Ã¢â‚¬ °Ã‚ ¤ F(Blackboard) 19:then Blackboard = Xk-means 20:reinitialize AFSA 21:Endif 22:until stopping criterion is met Figure (1): Pseudo code of proposed algorithm. The proposed algorithm along with standard AFSA, PSO algorithm, hybrid algorithm called KPSO [20], and k-means is used to segment two images, Lenna and Barbara. PSO and KPSO parameters are adjusted according to [6], and for k-means, initializing Forgy method is applied [21]. AFSA parameters and are adjusted according to [7]. AFSA settings in the proposed algorithm are the same as [7]. With respect to various experiments, if fitness value relating to Best AF is less than 0.1 in 3 iterations, it means that artificial fish swarm is converged. The following results are obtained from 50 times repeated experiments. Figure (3) shows segmented images, Lenna and Barbara, by the proposed algorithm with 5 and 3 thresholds. Figure 2: Orginal gray level Lenna (left) and Barbara (right) images Figure 3: The thresholded images of Lenna and Barbara using 5, and 2-level thresholds, from top to bottom. Average uniformity obtained from 5 algorithms on two images with thresholds 2, 3, 4 and 5 are shown in Table (1). As it is observed in Table (1), obtained results from the proposed algorithm is better than the other algorithms for all cases. AFSA algorithm has the worst result for all cases because of low ability in local search. K-means algorithm has found better results than AFSA because of high capability of k-means in local search. The reason for superiority of k-means to AFSA is the problem space property in histogram clustering. In fact, because of low dimensions of problem space in this environment, local search ability is of greater importance than global search ability. Also, it can reduce k-means weakness of sensitivity to initial values by means of one of the initializing methods of k-means like Forgy. Thereafter, with respect to considerable superiority of k-means local search ability in contrast to AFSA, k-means results are better than AFSAs. TABLE I: Comparison of uniformity for the five Algorithms Image T AFSA K-means PSO KPSO Proposed method Lenna 2 0.9138 0.9634 0.9730 0.9728 0.9775 3 0.9361 0.9749 0.9781 0.9783 0.9795 4 0.9495 0.9762 0.9816 0.9811 0.9826 5 0.9517 0.9804 0.9835 0.9834 0.9838 Barbara 2 0.9758 0.9761 0.9765 0.9768 0.9781 3 0.9783 0.9802 0.9808 0.9805 0.9820 4 0.9797 0.9834 0.9843 0.9851 0.9862 5 0.9822 0.9849 0.9855 0.9850 0.9884 Obtained results from PSO are better than k-means in all cases and its because of global search ability superiority of PSO to k-means. Moreover, in PSO, theres a trade-off between global search and local search abilities [16] and PSO also can perform a proper local search beside an acceptable global search. KPSO results are better than k-means results for all cases because after executing k-means in this algorithm, PSO algorithm is performed and improves obtained results from k-means. But obtained results from KPSO are not better than PSO for all cases. The reason is that sometimes k-means converges toward a local optimum and obtained result from that is not appropriate. Therefore, PSO is responsible for taking out the result from local optimum; however, it sometimes may not be successful. Indeed, improper result of k-means causes fast convergence of particles to local optimum. Obtained results from the proposed algorithm are better than other algorithms in all cases. The reason is u sage of strategies which have been used for global search in this algorithm. In fact, the proposed algorithm is successful in finding the global optima in most runs and can prevent final result from being trapped in local optima, whereas, this ability is observed less in other algorithms and they cannot guarantee passing local optima. This weakness causes that other algorithms to be of less robustness and not to be able to reach to almost the same results in their various implementations. Also, in the proposed algorithm, k-means algorithm performs local search after finding global optimum region by AFSA. Consequently, with respect to high ability of k-means in local search and taking proper initial cluster centers from AFSA, local search is done well in the proposed algorithm, too. As a result, both k-means and AFSA algorithms abilities are utilized in the proposed algorithm and the weakness of k- means algorithm cant decrease the algorithms efficiency. As it is observed in all algo rithms except KPSO, with rising up the number of thresholds, uniformity amount is improved. In KPSO, since the weakness of k-means has an undesirable effect on PSO efficiency, obtained results are not stable. In this paper, a new cooperative algorithm based on artificial fish swarm algorithm and k-means was proposed for image segmentation with respect to multi-level thresholding. In the proposed algorithm, AFSA performs global search and k-means is responsible for local search. The process of the proposed algorithm is such that the robustness and ability of preventing from being trapped in local optimums is improved. The proposed algorithm along with four other algorithms is used for segmenting 2 well-known images and obtained results are compared with each other. Experimental results show that segmented images quality by the proposed algorithm is much better than four other tested algorithms. [1] R. C. Gonzalez, and R. E. Woods, Digital image processing, In: Pearson Education India, Fifth Indian reprint, 2000. [2] S. Arora, J. Acharya, A. Verma., and K. Panigrahi, Multilevel thresholding for image segmentation through a fast statistical recursive algorithm, In: Journal on Pattern Recognition Letters 29, pp. 119125, 2008. [3] Maitra. M, A. Chatterjee, A hybrid cooperative-comprehensive learning based PSO algorithm for image segmentation using multilevel thresholding, In: Journal on Expert System with applications 34, pp. 1341-1350, 2008. [4] M. Mignote, Segmentation by fusion of histogram-based k-means clusters in different color spaces, In: IEEE Transactions on Image Processing, 2008. [5] X. Yang, W. Zhao, Y. Chen, and X. Fang, Image segmentation with a fuzzy clustering algorithm based on Ant-Tree, In: Journal of Signal Processing 88, pp. 2453-2462, 2008. [6] Y. T. Kao, E. Zahara, and I. W. Kao, A hybridized approach to data clustering, In: Journal on Expert System with Applications 34, pp. 1754-1762, 2008. [7] D. Yazdani, S. Golyari, and M. R. Meybodi, A new hybrid approach for data clustering, In: 5th International Symposium on Telecommunication (IST) , pp. 932937, Tehran, 2010. [8] J. Kennedy, and R. C. Eberhart, Particle swarm optimization, In: IEEE International Conference on Neural Networks, 4, pp. 1942 1948, Perth, 1995. [9] A. A. A. Esmin, D. L. Pereira, and F. Araujo, Study of different approach to clustering data by using the particle swarm optimization algorithm, In: IEEE Congress on Evolutionary Computation, pp. 18171822, Hong Kong, 2008. [10] L. X. Li, Z. J. Shao, and J. X. Qian, An optimizing method based on autonomous animate: fish swarm algorithm, In: Proceeding of System Engineering Theory and Practice, pp. 32-38, 2002. [11] D. Yazdani, S. Golyari, and M. R. Meybodi, A new hybrid algorithm for optimization based on artificial fish swarm algorithm and cellular learning automata, In: 5th International Symposium on Telecommunication (IST), pp. 932-937, Tehran, 2010. [12] D. Yazdani, A. N. Toosi, and M. R. Meybodi, Fuzzy adaptive artificial fish swarm algorithm, In: 23 th Australian Conference on Artificial Intelligent, pp. 334-343, Adelaide, 2010. [13] J. Hu, X. Zeng, and J. Xiao, Artificial fish swarm algorithm for function optimization, In: International Conference on Information Engineering and Computer Science, pp. 1-4, 2010. [14] Y. Luo, W. Wei, and S. X. Wang, The optimization of PID controller parameters based on an improved artificial fish swarm algorithm, In: 3rd International Workshop on Advanced Computational Intelligence, pp. 328-332, 2010. [15] C. X. Li, Z. Ying, S. JunTao, and S. J. Qing, Method of image segmentation based on fuzzy c-means clustering algorithm and artificial fish swarm algorithm, In: International Conference on Intelligent Computing and Integrated Systems (ICISS) , pp. 254- 257, Guilin, 2010. [16] L. Xiao, A clustering algorithm based on artificial fish school, In: 2nd International Conference on Computer Engineering and Technology, pp. 766-769, 2010. [17] D. Bing, and D. Wen, Scheduling arrival aircrafts on multi- runway based on an improved artificial fish swarm algorithm, In: International Conference on Computational and Information Sciences, pp. 499-502, 2010. [18] J. A. Hartigan, An overview of clustering algorithms, In: New York: John Wiley Sons , 1975. [19] C. Y. Tsai, and I. W. Kao, Particle swarm optimization with selective particle regeneration for data clustering, In: Journal of Expert Systems with Applications 38, pp. 65656576, 2011. [20] D. W. der Merwe, and A. P. Engelbrecht, Data clustering using particle swarm optimization, In: Congress on Evolutionary Computation, pp. 215-220, 2003. [21] E. Forgy, Cluster analysis of multivariate data: efficiency vs. interpretability of classification, In: Biometrics 21, pp. 768, 1965

Capital Punishment in the United States Essay -- Death Penalty Row Law

The death penalty is a controversial topic in the United States today and has been for a number of years. The death penalty is currently legal in 38 states and two federal jurisdictions (Winters 97). The death penalty statutes were overturned and then reinstated in the United States during the 1970's due to questions concerning its fairness (Flanders 50). The death penalty began to be reinstated slowly, but the rate of executions has increased during the 1990's (Winters103-107). There are a number of arguments in favor of the death penalty. Many death penalty proponents feel that the death penalty reduces crime because it deters people from committing murder if they know that they will receive the death penalty if they are caught. Others in favor of the death penalty feel that even if it doesn't deter others from committing crimes, it will eliminate repeat offenders. Death penalty opponents feel that the death penalty actually leads to an increase in crime because the death penalty desensitizes people to violence, and it sends the message that violence is a suitable way to resolve conflicts. Death penalty opponents also condemn the death penalty because of the possibility of an innocent person being put to death, and because it can be unfairly applied. Death penalty opponents feel that the death penalty must be abolished because it cheapens the value of human life. The death penalty desensitizes people to murder and violence because, by executing people, the state sends the message that violence is an acceptable means of resolving conflicts (Terrill). The death penalty also reduces the gravity of the loss of human life by making it legal for the state to kill people it deems to be beyond reform (Winters 57). Death penalty oppo... ...es, even though 80% of the population is in favor of it, because of the numerous ethical and practical issues that must be taken into consideration (Winters139-144). Experts on both sides of the argument have numerous statistics and studies to back up their claims and to refute the claims of their opponents. Death penalty supporters hold that the death penalty is a deterrent to crime, and brings justice to killers. However, death penalty opponents maintain that the death penalty does not deter criminals, and desensitizes people to violence. There are no easy answers to the questions surrounding the imposition of the death penalty in the United States. Thus one should pursue this question with an open mind and consider all sides of the argument, because as Thomas Jefferson once said, "difference of opinion leads to inquiry, and inquiry leads to truth" (Winters 11).

The Giver’s Compassion for Jonas :: The Giver Essays

The Giver’s Compassion for Jonas Jonas’ community is ordered and ruled. Everything is same: their clothes, houses and lives. People follow the rules until they die. They know nothing about the true human life. The receiver of memory, the giver, is the only person who is able to the true pleasure of life. When Jonas is elected as the receiver of memory by the community and meets the Giver, his life is changed. Everything he believes in was controlled and hidden the real human life by the community. He is getting to realize that he will not be able to stay in the community any more and starts to find his own and comfort place. I would like to focus on describing the Giver’ compression for Jonas because I do think that this book can not be described without him. In the book, the Giver is described as an old man, always staying and keeping his sadness for the community alone. He is the only person who really knows what is going on in the community and its people. His role is to give the community advice and help, when they face something that they have not experienced. The community needs the Giver, even thought they have a long history, already fixed its structure, rarely ask the Giver advice. â€Å"They know nothing,† The Giver said bitterly (p.105). He feels sad and helpless for the community, because they reject to have memory and choose painless and predictable life. Memory includes not only in sadness, pain, and evil of human life but also in real happiness and pleasure of human life. The Giver likes to have the memory, however he feels loneliness not to share the memory with people, regret to receive the memory, and bitterness that the community would keep this condition forever. He can not find out the possibility to chang e the entire community. Even though Jonas asked the Giver to come with him, he can not escape form the place and has to take care of the community (p.161-162). The Giver’s compassion in the story is not only for the community but also for Jonas. He must be confused and struggling because he already knows the train will be hard and control the most important parts of the human, which includes the feelings like love, warmth, sadness, patience, and pleasure.

The Problem of Moral Agency in Shakespeares Hamlet Essay -- GCSE Cour

The Problem of Moral Agency in Hamlet  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚     Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚   In order to be a moral agent, a person has to have a good sense of self, they have to know exactly who they are and how they must act according to the decisions they make. In Hamlet, the moral task at hand is revenge for the murder of Hamlet the elder. The murdered King's son, also of the same name, must be the one to avenge the murder. Before Prince Hamlet finds out the true story behind his father's death, he has his mother's "incestuous" remarriage to his uncle Claudius (who is now the King of Denmark) on his mind. Long after Hamlet learns the truth, he still does nothing. Hamlet is unable to act even though he has decided to seek revenge. One reason he does not act is because he cannot get past the fact that his mother is not, in his mind, adequately mourning old Hamlet's death. The second reason the Prince has problems with moral agency is because he does not really decide why he is planning to seek revenge on Claudius. His task is twofold, h e wants to avenge the murder of his father and he wants his mother to reveal her guilt about her hasty and incestuous marriage. Finally, Hamlet does not truly know who he is, and what he is to do until the very last act of Hamlet. This essay aims to explore why Prince Hamlet has trouble becoming a moral agent. When we first encounter Hamlet, his concerns are about his mother's remarriage to his uncle Claudius so soon after his father has died. The Prince is angry because Gertrude is not adequately mourning old Hamlet's death, and due to the insistence of Claudius that Hamlet consider him his father and king: O God, a beast that wants discourse of reason Would have mourn'd longer-- married with my uncle, My fathe... .... When Hamlet is doomed to die, he goes through with his revenge, but not for his father, nor for his mother-- The Prince finally kills the King when he finds out that it he, Claudius, who is responsible for the poisonous foil. This final reason to kill Claudius is most important of all. Works Cited Calderwood, James L.. To Be and Not To Be: Negation and Metadrama in Hamlet. --New York: Columbia University Press, 1983. Shakespeare, William. All's Well That Ends Well. --In: The Riverside Shakespeare. Ed. G. Blakemore Evans. --Boston: Houghton Mifflin Company, 1974; pp.504-541. Shakespeare, William. Hamlet. --In: The Riverside Shakespeare. Ed. G. Blakemore Evans. --Boston: Houghton Mifflin Company, 1974; pp. 1141-1186. Tirrell, Lynne. "Storytelling and Moral Agency."   --In: The Journal of Aesthetics and Art Criticism. --V.48, Spring 1990; pp.115-126.

TRW Case Analysis

Case 4 –TRW SYSTEMS (A and B condensed) ThiviyaManikandan sridhar – 54, Devika Srinivas – 11, Prabhudeep Shivakumar-31 1. What kinds of organizational design choices has TRW made about the design Challenges discussed in chapter 4? Due to the complexity of products being produced and the interdependency between the parts, systems, various groups, divisions and companies who assembled the parts forced TRW to adopt the matrix structure, where it covers vertical flow of functional responsibility and horizontal flow of product responsibility. On the vertical side, TRW systems have functional organizations like mechanical division, physical research division, systems divisions, fabrication integration and testing division. On the horizontal side, it has program organization which controls program office. Under these two organizations, sub project managers and assistant project managers are appointed. These managers have to report to two bosses. All these employees and departments are under the control of presidents and vice presidents. This is indicates a flat and decentralized structure, where managers and employees are allowed to take their own decisions. The level of decentralization followed by TRW is appropriate for the fluctuating and complex aerospace industry, as creativity plays important role in research and development which is accomplished by TRW engineers. A program manager maintained all the management responsibility for pulling together the various phases of a particular customer project. Assistant program manager was appointed to coordinate the activities of program manager. Under assistant program manager, sub project managers were also appointed to control the total project activity. Sub project managers are responsible for integrating and coordinating the functional organization and program organization. He also supervises the engineers and manages financial resources procured from program office. Sub project managers are the main integrating mechanism in the structure. There is too much of pressure and authority on the sub project managers, so TRW has to take drastic steps to develop its integrating mechanisms as integration plays an important role in matrix structure. TRW lacks standard operating procedures and standardized rules and norms. This indicates a total domination of mutual adjustment. As a result of this situation, most of the engineers are facing ambiguity problems. So TRW has to bring the balance between standardization and mutual adjustment. TRW makes minimal  use of formal hierarchical reporting relationships to coordinate activities. The informal  network of social relationships developed over time is important in determining how teams perform, and informal status  relationships between scientists are important as a means of coordination. Team values and norms  derive from informal interactions between scientists and are spread as members move between teams. 2. Are the design choices TRW has made appropriate for the organization, why or why not? From the above points, we can clearly say that the matrix structure and organic design followed by TRW systems is appropriate from the contingency perspective, as it matches the uncertain environment. But the managers are not utilizing their full potential and the employees are taking advantage of this structure by coming late to the work. This indicates the need for centralization and standardization. Fluctuations in the aerospace environment need spontaneous decisions which can be achieved by matrix structure. TRW is high in task variability and low in task analyzability. It  uses intensive technology and has  reciprocal interdependence. For all these features matrix structure is the best suitable structure. A matrix would not be suitable in  a simple, stable environment for routine technology and employees with  routine tasks. Here, it  would promote coordination and motivation problems and raise bureaucratic costs. 3. What is TRWs structure and what problems does it cause for TRW? TRW followed a matrix structure. The employees were responsible for two officers. Some of the problems caused by this structure – The relationship between the project officer and the functional division officer is a complex one. Both the roles are mutually dependent and have equal power. Hence authority is undefined. This leads to a lot of confusion about the role of the managers. Many employees are not comfortable with the relationship and this caused the status and authority problems within the organization. Another problem of the structure is the subproject manager is the prime mover of the organization. He is the person who brings the program officer’s requirements and the lab’s resources together to produce a subsystem. He has to cater to the needs of project manager as well as the functional manager. He has to keep in mind the interests of both the bosses; this puts him under a lot of pressure. If he reacts too much to pressure from either side, it hurts his ability to be objective about his subproject and this will in turn affect the employees. Hence the success of the project is majorly controlled by a single person. The matrix structure did not have any formal rules. The informal procedures followed are useful for the higher level management but the employees are given too much freedom. Due to the complexity of the structure, a lot of time is required in setting up a new project teams. This contributed to an increase in the cost incurred by the organization. As the structure is changing all the time there is lack of leadership. Also there exists large gap  between authority and responsibility. The project manager had no authority over people working on his project. He had to work with the functional heads on these problems. This imbalance enabled flexibility and adaptively in the organization, but it was difficult to work with. 4) What problems might TRW have with its present structure as it grows? The present matrix structure problem is that whether this matrix structure is suitable or Not for a large organization. As organization grows, it will be difficult for TRW to maintain Its Organic approach. We know that Divisions of TRW itself refused to share its R&D Information with other division. This attitude of employees will surely lead to Misunderstanding, conflicts and confusion. As the company grows, the company has to adopt For a new technology, some of the employees may not like new technology which is Complicated to understand and work.