** Assignment 1 **

** Creating machine learning model to predict customer churn for Telecomm company like Sprint **

** Steps to creating model **

1.Data Collection.

• Gather historical data on customers both who have stayed and those left on relevant features such demographics, usage patterns, contract details, customer support interactions, billing information, what they used to buy, etc.
• Preprocess the data, which involves handling missing values, encoding categorical variables, and scaling numerical features. Ensure data quality and consistency.

2.Data Exploration and Visualization.

• Understand the characteristics of the dataset.
• Identify trends, patterns, and correlations between features and churn.

3.Feature Engineering.

• Create features to improve the predictive power of the model. i.e. value and usage intensity for the last 1 year at an interval of 3 months.

4.Model Selection.

• Choose appropriate machine learning algorithms for classification i.e. logistic regression, decision trees, random forests, support vector machines, and gradient boosting algorithms like XGBoost or LightGBM.
• Consider using ensemble methods to combine the predictions of multiple models for improved accuracy.

5.Model Training.

Train the selected models on the training dataset.

==================================================================

** Assignment 2 **

** Measuring success of the Instagram TV product**

1.User Engagements through track of viewership, watch time,Likes, Comments, and Shares,Followers gained

2.Retention and User Behavior:

• How users regularly return to app, period spent and navigation through the app.

3.Monetization as to whether there are subscriber revenue channels and other channels like Ad Revenue

4.Demographics

• Content should be utilized across most countries.

5.Conducting user surveys and collect qualitative feedback to gauge user satisfaction, pain points, and feature requests related to IGTV.

6.Content Quality and Recommendations.

• Measure the effectiveness of the recommendation algorithm in suggesting IGTV videos that align with user interests and preferences.

7.Competitive Analysis.

Comparing IGTV's performance with competitors' platforms offering similar video content and ensuring to be outstanding

8.App Engagements to ensure there is more usage, longer sessions or more frequent logins.

9.Having Long-term Growth in terms of user base, content library, and revenue generation.

10.There should be profitability upon evaluate of operational and infrastructure costs associated with (IGTV).