Regression & Optimization
Budget: $30 – $250 CAD
Instructions:
The accompanying PDF file contains six questions, with an additional extra-credit component in Question 3. For these six questions, you will need to implement logistic regression from scratch mainly in Excel Solver, framing it as a maximum likelihood estimation problem. Additionally, based on the given price response curve, you will be required to optimize pricing using an alternative approach in price optimization.
At a high level, you will iteratively refine the maximum likelihood estimation and price optimization methodology, applying it to increasingly sophisticated models and custom segments.
Your answer should include a written analysis (in Word or PDF format) for each question, supplemented by relevant analytical components such as a spreadsheet model or other supporting materials. In your written analysis, ensure you clearly articulate the reasoning behind your solutions. Spreadsheet models should be well-annotated and structured to enhance readability and understanding.
You may also need to make a visualization likely on a per question basis as per below:
Q1 - Logistic Regression Model A scatter plot of historical bids with a logistic regression curve to show the probability of winning a bid as a function of price. A decision boundary plot showing the optimal price Fjord should offer.
Q2 - Segmented Analysis for Police vs. Corporate Buyers Two separate logistic regression curves for police vs. corporate buyers to visualize differences in price sensitivity. A bar chart or box plot comparing optimal prices and expected contributions for the two segments.
Q3 - Order Size Effect A 3D surface plot or heatmap showing how order size and price impact the probability of winning bids. A line chart showing optimal prices for different order sizes for police and corporate buyers.
Q4 - Manufacturing Cost Differences A cost vs. price comparison chart, highlighting the difference in manufacturing costs for police vs. corporate vehicles and how it affects pricing decisions.
Q5 - Model Improvements & Segmentation A tree diagram showing potential segmentation opportunities based on customer attributes.
Q6 - Fleet-Sales Process Optimization A flowchart illustrating the current and proposed fleet-sales processes to ensure Fjord captures optimal pricing benefits.
The accompanying PDF file contains six questions, with an additional extra-credit component in Question 3. For these six questions, you will need to implement logistic regression from scratch mainly in Excel Solver, framing it as a maximum likelihood estimation problem. Additionally, based on the given price response curve, you will be required to optimize pricing using an alternative approach in price optimization.
At a high level, you will iteratively refine the maximum likelihood estimation and price optimization methodology, applying it to increasingly sophisticated models and custom segments.
Your answer should include a written analysis (in Word or PDF format) for each question, supplemented by relevant analytical components such as a spreadsheet model or other supporting materials. In your written analysis, ensure you clearly articulate the reasoning behind your solutions. Spreadsheet models should be well-annotated and structured to enhance readability and understanding.
You may also need to make a visualization likely on a per question basis as per below:
Q1 - Logistic Regression Model A scatter plot of historical bids with a logistic regression curve to show the probability of winning a bid as a function of price. A decision boundary plot showing the optimal price Fjord should offer.
Q2 - Segmented Analysis for Police vs. Corporate Buyers Two separate logistic regression curves for police vs. corporate buyers to visualize differences in price sensitivity. A bar chart or box plot comparing optimal prices and expected contributions for the two segments.
Q3 - Order Size Effect A 3D surface plot or heatmap showing how order size and price impact the probability of winning bids. A line chart showing optimal prices for different order sizes for police and corporate buyers.
Q4 - Manufacturing Cost Differences A cost vs. price comparison chart, highlighting the difference in manufacturing costs for police vs. corporate vehicles and how it affects pricing decisions.
Q5 - Model Improvements & Segmentation A tree diagram showing potential segmentation opportunities based on customer attributes.
Q6 - Fleet-Sales Process Optimization A flowchart illustrating the current and proposed fleet-sales processes to ensure Fjord captures optimal pricing benefits.