Reference no: EM133174456

Problem Statement

BCCI has hired an external analytics consulting firm for data analytics. The major objective of this tie up is to extract actionable insights from the historical match data and make strategic changes to make India win. Primary objective is to create Machine Learning models which correctly predicts a win for the Indian Cricket Team. Once a model is developed then you have to extract actionable insights and recommendation.

Also, below are the details of the next 10 matches, India is going to play. You have to predict the result of the matches and if you are getting prediction as a Loss then suggest some changes and re-run your model again until you are getting Win as a prediction. You cannot use the same strategy in the entire series, because opponent will get to know your strategy and they can come with counter strategy. Hence for all the below 5 matches you have to suggest unique strategies to make India win. The suggestions should be in-line with the variables that have been mentioned in the given data set. Do consider the feasibility of the suggestions very carefully as well.

Question 1. 1 Test match with England in England. All the match are day matches. In England, it will be rainy season at the time to match.

Question 2. 2 T20 match with Australia in India. All the match are Day and Night matches. In India, it will be winter season at the time to match.

Question 3. 2 ODI match with Sri Lanka in India. All the match are Day and Night matches. In India, it will be winter season at the time to match.

Business Report: 1. Understand and define the problem statement.

2. Get a preliminary understanding of data and perform exploratory data analysis.

3. Discuss the business context.

4. Data cleaning and pre-processing (like outlier treatment, missing value treatment etc.)

5. How to generate insights from EDA?

6. Discuss any finer nuances that could be used to generate insights.

7. Discussion around what model performance measures could be applied?

8. Discuss model validation

9. Discuss model tuning

10. Discuss how to draw business insights & recommendations.

11. Business insights and recommendations

12. A structure for presentation.

13. How to further improve the model (if required).

14. Discuss what a good structure of ppt looks like and how to make a good business presentation.

15. A PPT that helps explain & communicate the data analysis to be used in the final presentation. Jupyter notebook files with coding of above business report sections majorly including as below:

1. Business Problem Understanding and Problem definition

2. Generate a data report.

3. Exploratory Data analysis and insights-driven from it.

4. Various Model Building and comparison

5. Model Tuning

6. Model Interpretation Coding should be briefed out for each models created (minimum 9-10 models) and referring to each perspective of the business requirement.

Attachment:- Problem Statement.rar