Discussion
TechGear Inc. Business Data Analysis
BDA Unit 6 Touchstone
ADD NAME HERE
ADD DATE HERE
Data Exploration and Summary (Question 1)
Analysis Task:
Import the data into Python and create a pandas DataFrame.
Summarize the structure of the data: the number of rows and columns.
Calculate the average advertising spend on Facebook and Instagram, and the average discount rate.
Slide Content:
Summary statistics (number of rows, columns, and averages).
Key insights from your initial data exploration.
Speaker Notes:
Describe the dataset and its key features.
Explain the importance of understanding the data before analysis.
Visualizing Relationships (Question 2)
Analysis Task:
Create scatter plots to visualize the relationships between sales and each of the following variables:
Facebook ad spend
Instagram ad spend
Discount rate
Slide Content:
Summary statistics (number of rows, columns, and averages).
Key insights from your initial data exploration.
Speaker Notes:
Describe the dataset and its key features.
Explain the importance of understanding the data before analysis.
Simple Linear Regression (Question 3)
Analysis Task:
Develop a simple linear regression model to predict sales based on Facebook ad spend.
Interpret the slope and R-squared value.
Slide Content:
Regression output from Python.
Summary of your interpretation of the model’s coefficients and R-squared value.
Speaker Notes:
Explain what the slope indicates about the relationship between Facebook ad spend and sales.
Discuss the R-squared value and what it tells you about the model’s fit.
Assessing Model Fit (Question 4)
Analysis Task:
Assess the fit of your simple linear regression model by analyzing the residuals.
Create a residuals vs. fitted values plot and a Q-Q plot.
Slide Content:
Residuals plots and summary of their implications for model reliability.
Speaker Notes:
Explain what the plots reveal about the accuracy and reliability of the model.
Multiple Linear Regression (Question 5)
Analysis Task:
Develop a multiple linear regression model using Facebook ad spend, Instagram ad spend, and discount rate to predict sales.
Compare this model’s performance to the simple linear regression model.
Slide Content:
Regression output and performance comparison.
Key insights about how these variables collectively influence sales.
Speaker Notes:
Interpret the coefficients of the multiple linear regression model.
Discuss how the model improves on the simple linear regression model.
Forecasting (Question 6)
Analysis Task:
Construct both a 3-month moving average forecast and an exponential smoothing forecast for January 2025 (use a smoothing parameter of 0.80).
Choose the most reliable method based on TechGear’s preference for recent sales trends.
Slide Content:
Forecast results from both methods.
Recommendation of the most reliable method.
Speaker Notes:
Justify your choice of forecasting method.
Provide recommendations on how TechGear can use the forecast to optimize marketing strategies.
Machine Learning Models (Question 7)
Analysis Task:
Build a multiple linear regression model using 5-fold cross-validation to predict future sales.
Build a decision tree model using 5-fold cross-validation to predict future sales.
Compare the performance of these models using RMSE and determine which one TechGear should select. Provide a reason for why TechGear should select the model you are suggesting.
Slide Content:
RMSE comparison and model selection.
Key insights from the best model.
Speaker Notes:
Explain why you selected the best model.
Interpret the RMSE and discuss its implications for decision-making.
Consider the $6,500 threshold and the potential range of actual sales.
Monte Carlo Simulation (Question 8)
Analysis Task:
Estimate the average and median monthly sales by running 1,000 simulations.
Assume that daily sales follow a uniform distribution within the minimum and maximum sales values TechGear experienced over the last 60 months.
Use a random seed value of 42.
Interpret the standard deviation of the simulated monthly sales and explain its significance in understanding the variability of TechGear's sales.
Slide Content:
Histogram and summary statistics (average, median, and standard deviation).
Key insights on sales variability.
Speaker Notes:
Explain the significance of the average, median, and standard deviation.
Provide actionable insights for improving sales, budgeting, and operations based on the simulation.
Linear Programming (Question 9)
Analysis Task:
Use linear programming to optimize TechGear’s monthly advertising budget allocation between Facebook and Instagram to maximize sales.
Ensure you meet all budget and constraint requirements.
Slide Content:
Optimal budget allocation and maximum sales result.
Speaker Notes:
Explain the optimal allocation between Facebook and Instagram advertising.
Describe the maximum achievable sales and how this allocation aligns with business goals.