PRDA-01
PROFIT ANALYSIS
DONE BY
VINCYPAUL.F
PTID-CDA-DEC-24-265
15-JUL-24-CDA-BUN-040-
WDE20-ONL
INTRODUCTION TO THE PROJECT AND ITS
OBJECTIVE
• Our project aims to analyze a dataset containing
information about companies' financial attributes,
including R&D spending, Administration spending,
Marketing spending, State, and Profit.
• Retrieve data from the database.
• Perform regression analysis to evaluate the impact of R&D, Marketing, and
Administration spend on profit.
• Predict profit for given input features.
• Visualize data and provide actionable insights.
PROFIT ESTIMATION OF A COMPANY
STEP 1: DATA PREPARATION AND
ANALYSIS
• We loaded the dataset using the provided link and credentials.
• The dataset was loaded using pandas from a CSV file.
• We displayed the entire dataset, checked for missing values, and generated
summary statistics for all columns.
• Descriptive statistics and visualizations are performed to get an initial
understanding of the data.
• The output confirmed that there were no missing values, and we obtained
key statistics such as mean, standard deviation, and quartiles for numeric
columns.
PYTHON CODE COULD BE ADAPTED TO CREATE A SOLUTION
FOR STEP 1:
OUTPUT
STEP 2: REGRESSION ANALYSIS
• Choose the appropriate regression model (e.g., multiple linear regression) to predict profit based on
R&D spending, Administration spending, and Marketing spending.
• Split the data into training and testing sets.
• We split the dataset into features (R&D spending, Administration, Marketing spending) and the target
variable (Profit).
• Train the regression model on the training data.
• Evaluate the model's performance on the testing data using metrics like R-squared, Mean Absolute
Error (MAE), etc.
• The data was divided into training and testing sets using the train_test_split function.
• We created a Linear Regression model, trained it, made predictions, and evaluated its performance
using Mean Squared Error (MSE) and R-squared (R2) scores.
• The R2 score indicated that the initial model did not fit the data well and requires improvement.
PYTHON CODE COULD BE ADAPTED TO CREATE A SOLUTION FOR STEP 2:
STEP3:PREDICT PROFIT AND OPTIMIZATION
• Use the trained regression model to predict profit based on input
features (R&D spending, Administration spending, Marketing
spending).
• Use Solver or another optimization technique to find the optimal
spending on R&D, Administration, and Marketing that maximizes
profit.
• We performed constrained optimization to maximize profit
coefficients for R&D, Administration, and Marketing spending.
N
• The optimization process successfully determined the coefficients for the features. Here are the
optimized coefficients:
• R&D Coefficient: 0.7098
• Administration Coefficient: 0.2473
• Marketing Coefficient: 0.0429
• Insights:
• R&D spending has the highest influence on profits, indicating it is the most impactful investment
area.
• Administration has a moderate contribution, showing it plays a supportive role in profitability.
• Marketing spending has a relatively low impact, suggesting the company might optimize or
reconsider its current marketing strategies.
ACTUALVSPREDICTPROFIT
Marketing
R&D Spending Administration Actual Profit Predicted Profit
Spending
165349.20 136897.80 471784.10 192261.83 190705.10
162597.70 151377.59 443898.53 191792.06 189775.72
153441.51 101145.55 407934.54 191050.39 176703.75
144372.41 118671.85 383199.62 182902.00 171104.72
142107.34 91391.77 366168.42 166187.94 166187.94
OBSERVATIONS:
1.The predicted profits closely align with the actual profits for most entries, indicating the model is performing well.
Predict the Profit for the below-given input features:
Marketing
R&D Spending Administration Predicted Profit
Spending
21892.92 81910.77 164270.7 67,308.11
23940.93 96489.63 137001.1 70,217.62
STEP 4: DATA VISUALIZATION AND
INSIGHTS
• Created visualizations using Tableau to represent relationships
between different features and profit.
• A scatter plot was generated to visualize the relationship between
R&D Spending and Profit.
• A BARCHART for numerical variables.
• A correlation heatmap illustrated the correlations between R&D
Spending, Administration, Marketing Spending, and Profit.
• The strongest correlation was identified between R&D Spending
and Profit.
DASHBOARD VISUALIZATION OF SCATTERPLOT,HEATMAP,BARCHART
SPECIFIC INSIGHTS FOR THE
PROVIDED DATA
• Heatmap: New York shows the highest profit contribution;
targeted strategies could amplify this success.
• Scatter Plot: R&D spending has the strongest positive
correlation with profit, suggesting prioritization in budget
allocation.
• Bar Chart: Comparison of actual vs. predicted profits
highlights model accuracy and areas for refinement.
CONCLUSION
• In conclusion, our project involved comprehensive data preprocessing, regression
analysis, optimization, and insightful data visualizations. Visualizations provided
valuable insights into the dataset's attributes and relationships.
• R&D spending has the highest impact on profit.
•Marketing and administration need targeted optimization.
•State-level strategies can unlock new growth opportunities.
•Adopting these measures will enhance profitability and competitiveness.
RECORDED AUDIO PRESENTATION
THANK YOU
