Discussion on Responsible Management

 Question 1

“What does ‘responsible management’ mean to you? Take a photo symbolizing an aspect of ‘responsible management’ and write a short abstract (maximum 200 words) about why this photo symbolizes responsible management in your view.”

The environment consists of different components which depend on each other for optimal performance and survival. When any factory, company, processing plant or any institutions operates from a particular locality it is integral to care for the surroundings. Any organization in an area should strive to ensure that despite its revenue generation ventures in an area it is important to ensure that they at least compensate for the negative impacts on the environment. The organization should also strive to ensure that its negative impact on the environment is minimal.

///////////////

The photo above shows petroleum processing plant employees taking part in efforts to clean the environment. Petroleum processing plants emit a lot of waste into the environment. The waste is mainly is gaseous, solid or liquid. Some of this waste finds itself in the ocean where it is located. Waste from petroleum processing has a negative impact on organisms that live in water and affects economic activities such as tourism and fishing. The company sets out a team from time to time to clean beaches in an effort for responsible management. These ensure that the negative impact that the plant has on the environment is reduced. These operations enable the plant to earn the trust of the surrounding community and the government.

Question 2

“Illustrate the ‘ETL’ process, using a dataset that was not analysed before.”

The Extract Load Transform (ETL) process entail the extraction of the data various sources and its transformation. Loading of the data into warehouses after the two steps is done. To illustrate the ETL process we analyze records from the University of Wisconsin on breast cancer patients. Step number one in the ETL process is extracting data from where it is being sourced. Before the physical extraction and loading of the data, it is important to have a logical data map. For breast cancer patients data set the extraction mode used is the full extraction method. Validation of the data through reconciling the data with its source should be performed. The data is processed to eliminate duplicates and ensure that all keys in the data are in place. The data in the extract stage is raw and unusable.

Step number two in the ETL process transforms the data. To be useful, the data needs to be cleaned, mapped and some transformations applied to it. In this stage, it is possible to customize data operations. Data integrity problems such as cases where one person’s name has different spellings. During this stage, the data is also validated. The data is validated through activities such as filtering of the data set to select the suitable columns for use. The original data set has over 24 columns each carrying a certain attribute of each patient. Only ten columns are selected from this data set. The attributes are indexed 2 through to 10. These instances are typical of instance-based machine learning.

After selecting the columns for use the data is cleaned. Each patient has two possible conditions. These conditions are either malignant or benign. To identify different patterns in the data pattern separation is performed on the data. At the time of performing pattern separation, there were 369 instances in the data. Only one trail results from the classification of the data. Consistency with two pair arranged hyperplanes is shown by half the data. The other half produces a 94% accuracy. Consistency with three pair arranged parallel planes were found in two-thirds of the data. The remaining third of the data produces 96% accuracy. Selection of data points suitable for machine learning is selected.  Applying a machine learning algorithms based on four instances on the data resulting from classification results in dropping of two instances from the 369 instance data set. Entries which were dropped are 765878 and 484201. Zero bare nuclei are replaced in instance 1080185 and 1187805. Field 6 has changed from 0 to 1 in instance 1219406. The classifying attribute is placed in the last column.  

The ETL process’ final step is the load stage. Data is loaded into warehouses. In typical cases, large data sets need to be loaded into the warehouse in short periods necessitating the optimization of the load process to improve performance of the ETL process. The data warehouse automatically updates as the data changes. The data set for the patients at Wlicocsin adds up to 698 instances and 12 attributes. Of the 11 attributes, there is a class attribute.

Question 3

 “Use one supervised and one unsupervised approach to analyze a data set”

Models have to be trained and tested to determine their performance. Training and testing are done in both supervised and unsupervised learning. Training and testing the data requires the presence of two data sets from the same population. Using the same data during the training and testing of the model results in an incorrect analysis of the model (Chilimbi et al 2014). Same data set in training and testing results in a 100% accurate model. The data set is split into two. Three-quarters of the data is used in training the model while the remaining quarter is used in model performance testing. How well a model fits the data set is used to determine its accuracy. High variability in data being analyzed is a downside. Same levels of magnitude are therefore essential in data. To create the same levels of magnitude, the transformation of data to fit a scale of 1-200 is done.

To make use of the supervised approach the data must have labels to guide the algorithm. The cancer patients’ data set from Wisconsin has ten variables with one dependent and nine independent. Mapping an input to output based on previous input-output pairs’ examples is known as supervised learning (Rusell et al, 2010). Splitting of the data provides both input and output for the supervise analysis approach. The cancer patients’ data set is categorical and it, therefore, contains values and their labels. Conversion of labels into numeric values is done i.e Malignant(M) is assigned 1 while the benign (B) is assigned 0.  The two main categories of supervised classification are classification and. When a data set contains continuous variables regression can be used for modelling the data (Kapelner and Bleich, 2013). Applying a filter to the data is known as classification.

Unsupervised models are different from supervised models in that they can run on their own without supervision (Kassambara, 2017). No labels or classification are required for analysis using the unsupervised approach. The unsupervised learning algorithm does not need any guidance when acting on the data. Acting on the data without guidance allows for unknown pattern discoveries (Scalettar et al, 2017). Categorization makes use of the unsupervised approach in finding common features. The real-time analysis in unsupervised learning allows data analysis and labelling.

The breast cancer data set in this study has nine independent variables and one dependent variable. The dependent variable takes the value 1 and 0 representing Malignant and Bening respectively. The best model for the analysis of the data is the classification algorithm which is under a supervised approach. Different classification methods including logistic regression, SVM and Naïve Bayes algorithm are applied to the data.

Analysis of data sets makes use of the clustering technique in helping find patterns in data which is uncategorized (Loh, 2011). Applying clusters on the data results in two natural clusters 1 for malignant and 0 for benign. Data is grouped such that it can either belong to 0 or 1 through explosive partitioning.   Another cluster known as the agglomerative cluster, which is a type of k-means, is used on the data. Each instance is formed as a cluster. Merging of clusters is done to decrease the count of clusters in the data.

Question 4

Visualize the results from your analysis for both supervised and unsupervised approach) in an appropriate way. Justify your choices

Visualization is important in analysis. It is advisable for an analyst to know what is in the data and what is not.

Figure 1 The last five instances of the data set

The data has M representing malignant and B representing being in the diagnosis column which is the dependent variable. A summary of the data based on the diagnosis attribute yields 357 instances for B and 212 for M.

Figure 2 visualization of the data set analysis

Figure 3 Visualization of the analysis of the data set’

The analysis is performed using the supervised approach. Through inbuilt python classed it is possible to tell when a classification algorithm makes a mistake. The classification performed on the data produces the following accuracy results i.e when the performance of the model is tested. All models fair relatively well exceeding 90%.

  1. SVM – 97.8%
  2. Decision Trees algorithm – 96%
  3. Logistic regression – 96.2%
  4. Kernel SVM – 97.9%
  5. K nearest neighbour – 95%
  6. Naive Bayes- 91.6%

With an accuracy of 91.6%, the Naïve Bayes model performs the poorest. The best performing model is the Kernel support vector machine followed by the Support vector machine. When the kernel support vector machine is fed on the data for analysis it produces an accuracy of 97%.

REFERENCES

Criminisi, A., Shotton, J., & Konukoglu, E. (2012). Decision forests: A unified framework for classification, regression, density estimation, manifold learning and semi-supervised learning. Foundations and Trends® in Computer Graphics and Vision7(2–3), 81-227.

 Stuart J. Russell, Peter Norvig (2010) Artificial Intelligence: A Modern Approach, Third Edition, Prentice-Hall ISBN 9780136042594.

Loh, W. Y. (2011). Classification and regression trees. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery1(1), 14-23.

Durden, C. (2008). Towards a socially responsible management control system. Accounting, Auditing &     Accountability Journal.

Kapelner, A., & Bleich, J. (2013). bartMachine: Machine learning with Bayesian additive regression trees. arXiv preprint arXiv:1312.2171.

Kassambara, A. (2017). A practical guide to cluster analysis in R: Unsupervised machine learning (Vol. 1). STHDA.

Hu, W., Singh, R. R., & Scalettar, R. T. (2017). Discovering phases, phase transitions, and crossovers through unsupervised machine learning: A critical examination. Physical Review E95(6), 062122.

Chilimbi, T., Suzue, Y., Apacible, J., & Kalyanaraman, K. (2014). Project adam: Building an efficient and scalable deep learning training system. In 11th {USENIX} Symposium on Operating Systems Design and Implementation ({OSDI} 14) (pp. 571-582).

Rahm, E., & Do, H. H. (2000). Data cleaning: Problems and current approaches. IEEE Data Eng. Bull.23(4), 3-13.

Hu, W., Singh, R. R., & Scalettar, R. T. (2017). Discovering phases, phase transitions, and crossovers through unsupervised machine learning: A critical examination. Physical Review E95(6), 062122.

Appendix 1

Figure 4The ETL process

Appendix 2

Figure 5 Data set after importing

Get professional assignment help cheaply

Are you busy and do not have time to handle your assignment? Are you scared that your paper will not make the grade? Do you have responsibilities that may hinder you from turning in your assignment on time? Are you tired and can barely handle your assignment? Are your grades inconsistent?

Whichever your reason may is, it is valid! You can get professional academic help from our service at affordable rates. We have a team of professional academic writers who can handle all your assignments.

Our essay writers are graduates with diplomas, bachelor, masters, Ph.D., and doctorate degrees in various subjects. The minimum requirement to be an essay writer with our essay writing service is to have a college diploma. When assigning your order, we match the paper subject with the area of specialization of the writer.

Why choose our academic writing service?

  • Plagiarism free papers
  • Timely delivery
  • Any deadline
  • Skilled, Experienced Native English Writers
  • Subject-relevant academic writer
  • Adherence to paper instructions
  • Ability to tackle bulk assignments
  • Reasonable prices
  • 24/7 Customer Support
  • Get superb grades consistently

Get Professional Assignment Help Cheaply

fast coursework help

Are you busy and do not have time to handle your assignment? Are you scared that your paper will not make the grade? Do you have responsibilities that may hinder you from turning in your assignment on time? Are you tired and can barely handle your assignment? Are your grades inconsistent?

Whichever your reason may is, it is valid! You can get professional academic help from our service at affordable rates. We have a team of professional academic writers who can handle all your assignments.

Our essay writers are graduates with diplomas, bachelor's, masters, Ph.D., and doctorate degrees in various subjects. The minimum requirement to be an essay writer with our essay writing service is to have a college diploma. When assigning your order, we match the paper subject with the area of specialization of the writer.

Why Choose Our Academic Writing Service?

  • Plagiarism free papers
  • Timely delivery
  • Any deadline
  • Skilled, Experienced Native English Writers
  • Subject-relevant academic writer
  • Adherence to paper instructions
  • Ability to tackle bulk assignments
  • Reasonable prices
  • 24/7 Customer Support
  • Get superb grades consistently

How It Works

1.      Place an order

You fill all the paper instructions in the order form. Make sure you include all the helpful materials so that our academic writers can deliver the perfect paper. It will also help to eliminate unnecessary revisions.

2.      Pay for the order

Proceed to pay for the paper so that it can be assigned to one of our expert academic writers. The paper subject is matched with the writer’s area of specialization.

3.      Track the progress

You communicate with the writer and know about the progress of the paper. The client can ask the writer for drafts of the paper. The client can upload extra material and include additional instructions from the lecturer. Receive a paper.

4.      Download the paper

The paper is sent to your email and uploaded to your personal account. You also get a plagiarism report attached to your paper.

 

smile and order essaysmile and order essayPLACE THIS ORDER OR A SIMILAR ORDER WITH MY ESSAY HERO TODAY AND GET AN AMAZING DISCOUNT

order custom essay paper

Leave a comment

Your email address will not be published. Required fields are marked *