Reference no: EM133870519 , Length: word count:1500

Assessment - Benchmarking

Purpose: The purpose of this assessment is to assess the student's ability to perform a benchmarking exercise on the chosen methodology/model in Assessment 3, and optimise the existing model and compare with other models.

Topic: Benchmarking, optimisation

Task Details:
Students will conduct a benchmarking exercise on the same dataset they have used in Assessment. Individually, students will re-visit the analysis and methodology they have used in Assessment 3 and perform different types of benchmarking and optimisation on the methodology.

First, students will determine different parameters of the chosen Predictive model from Assessment 3, research on these parameter optimsation process, and apply/perform optimisation of the parameters to improve different performance matrix of the chosen Predictive Model from Assessment 3. Second, students should then choose a few other Predictive Model techniques and apply it on their chosen datasets and assess improvements (if any) of different performance matrix of the alternative modeling techniques. For example, if students have chosen Supervised Classification technique for their Assessment 3 and have used Decision Tree model for classification, they could first investigate several parameters of the Decision Tree model and research the parameters and optimise those parameters to observe any improvements of different performance matrix could be made. Then, students can identify alternative Classification techniques, such as Random Fo;cest, Naïve Bayes, SVM etc. and utilise those techniques on the same problem dataset/s and evaluate different performance matrix and compare against Decision Tree Model. Students may also perform similar parameter optimisation process for the alternative Predictive Model algorithms and evaluate the results.

Finally, students have to write a 1500-word report outlining the detail of different machine learning techniques used, the parameter optimisation process of each of the prediction model techniques they have investigated and evaluated, and the different performance matrix they have assessed for each of these techniques. Students, have to provide their analytical thoughts on the comparison, explain why the new results are better / worse than the previous results and finalise their recommendation to the business/enterprise for the leadership decision making for the chosen industry/enterprise based on their comprehensive finding through Assessment 2, 3 and 4.