Business Problem:

· Popularity of SUV segment in the market

Classification Models Implemented:

· Logistic Regression

· K-Nearest Neighbors (K-NN)

· Support Vector Machine

Model Insights:

· Identified the potential customers for upgrade

· Identified factors predominantly influencing

Business Problem:

· Prioritization of budget allocation to departments based on the extent to which they influence the profitability

Regression Models Implemented:

· Multiple Linear Regression

· Decision Tree Regression

· Random Forest Regression

Model Insights:

· Identified the most influencial departments on the profitability

· Optimised budget allocation plan to maximise the profitability

Project Description

National Grid is one of the largest investor-owned energy companies in the world covering UK and north eastern US such as Massachusetts, New Hampshire, Rhode Island, New York (upstate, New York City and Long Island).

Time Series Analysis: Gas Consumption

National Grid want to estimate the gas consumption for future years based on region for domestic, commercial and industrial usage.

Business Problem:

· Evaluate the Daily Gas Active Storage field’s capacity for future demand based on region

Time Series Models Implemented:

· ARIMA

· SARIMAX

· Exponential Smoothing

Business Insights:

· More accurate estimation of future demand

· Gas storage capacity infrastructure design (Active Storage fields)

· Procurement and optimal inventory management

Regression Analysis:

Business Problem:

· Estimate monthly average Gas consumption volume per customer.

Regression Models Implemented:

· Multiple Linear Regression

· Decision Tree Regression

· Random Forest Regression

Business Insights:

· Utilization pattern at customer level

· Monthly based planning

· More accurate planning in high consumption seasons

· Forecast demand based on location

Business Problem:

· Knowing the customer’s satisfactory levels

Models Implemented:

· Bag of words – Logistic Regression

· Bag of words – Decision Tree Regression

Model Insights:

· Identified the areas to improve, preferred choice dishes