Automobile Sales Data Analysis and Prediction

Automobile Sales Data Analysis and Prediction/Dataset/README.md

@@ -0,0 +1,3 @@
+The dataset contains Sales data of an Automobile company.
+
+https://www.kaggle.com/datasets/ddosad/auto-sales-data/data

Automobile Sales Data Analysis and Prediction/Model/README.md

@@ -0,0 +1,101 @@
+## **Automobile Sales Data Analysis and Prediction**
+
+### 🎯 **Goal**
+
+The aim of this project is to create a machine learning model to predict the sales of the automobiles and prepare a data analysis of the same.
+
+### 🧵 **Dataset**
+
+https://www.kaggle.com/datasets/ddosad/auto-sales-data
+
+### 🧮 **What I had done!**
+
+1. Data Cleaning and Preprocessing
+
+- **Load the dataset**: Import the data into a pandas DataFrame for manipulation and analysis.
+- **Convert date columns to datetime format**: Ensure date columns are in datetime format for time series analysis.
+- **Handle missing values**: Impute missing numerical values with mean/median and categorical values with mode to maintain dataset integrity.
+- **Ensure appropriate data formats**: Verify that all columns have the correct data types (e.g., integers, floats, strings).
+- **Remove duplicates**: Identify and eliminate duplicate rows to avoid skewed analysis results.
+
+2. Exploratory Data Analysis (EDA)
+
+- **Summarize the dataset with descriptive statistics**: Get an overview of the data distribution using mean, median, and standard deviation.
+- **Plot distributions**: Use histograms and boxplots to visualize the distribution of numerical variables and identify outliers.
+- **Analyze relationships**: Create a correlation matrix to examine relationships between variables and use scatter and bar plots to explore these relationships further.
+
+3. Feature Selection and Engineering
+
+- **Identify relevant features**: Use domain knowledge and statistical methods (like correlation) to select features that impact sales.
+- **Create new features if needed**: Develop aggregate or interaction features that could provide additional insights.
+- **Encode categorical variables**: Apply One-Hot Encoding for nominal categories and Label Encoding for ordinal categories to convert them into numerical format.
+- **Scale numerical features**: Use StandardScaler or MinMaxScaler to normalize the numerical data, ensuring consistent input ranges for models.
+
+4. Model Selection and Training
+
+- **Split data into training and testing sets**: Divide the dataset into training (e.g., 80%) and testing (e.g., 20%) subsets to evaluate model performance.
+- **Choose regression algorithms**: Select models like Random Forest Regressor and Gradient Boosting Regressor for their effectiveness in handling regression tasks.
+- **Train models and perform hyperparameter tuning**: Fit the models on the training data and use cross-validation or grid search for hyperparameter optimization.
+
+5. Model Evaluation
+
+- **Evaluate performance using metrics**: Assess models using Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and R-squared (R²) to understand their accuracy.
+- **Compare model performance**: Analyze and compare the evaluation metrics to select the best-performing model.
+
+6. Interpretation and Insights
+
+- **Extract insights from model results**: Use feature importance scores and partial dependence plots to understand which features most influence sales.
+- **Analyze the impact of features on sales**: Draw conclusions about how different car features and economic conditions affect sales performance.
+
+7. Visualization
+
+- **Visualize actual vs. predicted sales**: Create line plots for time series data and scatter plots for predicted vs. actual sales values to assess model accuracy.
+- **Use bar plots and residual plots**: Display feature importances with bar plots and check model accuracy with residual plots, highlighting areas for improvement.
+
+### 🚀 **Models Implemented**
+
+1. Simple Dense Model
+2. Bidirectional LSTM
+3. Convolutional Neural Network
+4. Random Forest Regressor
+5. Temporal Convolutional Network
+
+### 📚 **Libraries Needed**
+
+1. numpy
+2. pandas
+3. matplotlib
+4. scikit-learn
+
+### 📊 **Exploratory Data Analysis Results**
+
+<img src = "https://github.com/why-aditi/ML-Crate/blob/main/Automobile%20Sales%20Data%20Analysis%20and%20Prediction/Images/Dealsize_bar.png"/>
+<img src = "https://github.com/why-aditi/ML-Crate/blob/main/Automobile%20Sales%20Data%20Analysis%20and%20Prediction/Images/Dealsize_pie.png"/>
+<img src = "https://github.com/why-aditi/ML-Crate/blob/main/Automobile%20Sales%20Data%20Analysis%20and%20Prediction/Images/Productline_bar.png"/>
+<img src = "https://github.com/why-aditi/ML-Crate/blob/main/Automobile%20Sales%20Data%20Analysis%20and%20Prediction/Images/Productline_pie.png"/>
+<img src = "https://github.com/why-aditi/ML-Crate/blob/main/Automobile%20Sales%20Data%20Analysis%20and%20Prediction/Images/Status_bar.png"/>
+<img src = "https://github.com/why-aditi/ML-Crate/blob/main/Automobile%20Sales%20Data%20Analysis%20and%20Prediction/Images/Status_pie.png"/>
+
+### 📈 **Performance of the Models based on the Accuracy Scores**
+
+`MAE was used as the performance metric.
+
+1. Simple Dense model: 575.88
+2. Bidirectional LSTM: 9.149645e+02
+3. Convolutional Neural Network: 9.461755e+02
+4. Random Forest Regressor: 610.215271
+5. Temporal Convolutional Network: 698.471619`
+
+### 📢 **Conclusion**
+
+`**Best Performing Model**: Model 1 now has the lowest MAE (575.88), MSE (551668.13), RMSE (742.74), and MAPE (20.27), indicating that it has the lowest overall errors among the five models.
+
+**Second Best**: Model 4, which previously was the best performer, now comes in second with slightly higher MAE (610.22), MSE (610762.19), RMSE (781.51), and MAPE (22.02) compared to Model 1.
+
+**Models 2 and 3**: These models continue to have significantly higher errors across all metrics, making them the least preferred models.
+
+**Model 5**: This model performs moderately but not as well as Models 1 and 4. Its errors are higher in all metrics compared to Model 1 and Model 4.`
+
+### ✒️ **Your Signature**
+
+`Aditi Kala`

Automobile Sales Data Analysis and Prediction/requirements.txt

@@ -0,0 +1,4 @@
+numpy==1.26.4
+pandas==2.2.2
+matplotlib==3.9.0
+scikit-learn~=1.5.0