Utilizing Machine Learning Techniques in Inventory Planning and Control of Automotive Spare Parts Table of contents Graduation Project 1 Graduation Project 2 Data Collection Data Preprocessing Basic Calculation Moving Average Artificial Neural Network Demand Prediction using Artificial Neural Network Autoregressive Integrated Moving Average Demand Prediction using Autoregressive Integrated Moving Average Comparative Analysis Constraints and Challenges Conclusion Introduction Problem Statement Scope Problem Statement Objectives Machine Learning Models Literature Review ANN ARIMA SVM Data Collection Item Car Perfume 001 Car Care Products 002 Car Care Products 003 Car Products 004 Car Care Products 005 Car Care Products 006 Bulb 12 V 007 Basic Calculations Basic Calculations Basic Calculations Dataset Preprocessing Car Perfume 001 Car Care Products 002 Car Care Products 003 Car Perfume 004 Car Care Products 005 Car Care Product 006 Bulb 12 V 007 Moving Average Moving Average Artificial Neural Network (ANN) Mathematical Background Artificial Neural Network Structure Artificial Neural Network Structure Activation Function Training algorithm Backpropagation adjusts weights. Architecture Layers connect input to output. Demand Prediction using ANN Time series Selected Car Perfume 001. Used a Nonlinear Autoregressive Neural Network (NAR) for time series forecasting, predicting sales quantities based on past data. 80% Training , 20% Validation and Testing 19 Network Diagram Model Development 20 Training Curve Performance 21 Error Histogram 22 Training, Testing, and Validation 23 Training set 24 Performance Metrics 25 Actual VS Predicted Sold Quantities 26 Training Algorithms and Structure ARIMA (Autoregressive Integrated Moving Average) Algorithm ARIMA Mathematical Background Combining the AR, I, and MA components, the ARIMA (p, d, q) model is given by:   Demand Prediction for ARIMA 1 A 95% confidence interval SPSS software Expert Modeler tool was employed Model Description The ARIMA parameters p,d,q The input time series is represented by Model ID as independent variable The Sold Quantity as dependent variable Traning and learning Analyzing Residuals using Autocorrelation Function (ACF) and Partial Autocorrelation (PACF) Performance Metrics Root Mean Squared Error(RMSE): evaluates prediction accuracy Coefficient of Determination (R2 ): indicates how well the regression model approximates the real data points Model Parameters Actual VS Predicted Sold Quantities Actual VS Predicted Sold Quantities Actual VS Predicted Sold Quantities Performance Metrics Analyzing Residuals using Autocorrelation Function (ACF) and Partial Autocorrelation (PACF) Comparative Analysis of the Developed Prediction Models Constraints and Challenge General Constraints Project-Specific Constraints Conclusion media1.mp4 media2.mp4 image1.jpeg image2.jpeg image3.png image4.png image5.png image6.png image7.png image8.jpeg image9.jpeg image10.jpeg image11.jpeg image12.jpeg image13.jpeg image14.jpeg image15.jpeg image16.jpeg image23.png image24.png image17.png image18.png image19.png image20.png image21.png image22.png image25.png image26.png image27.png image28.png image29.png image30.png image31.png image32.png image33.png image34.png image35.jpg image36.png image37.png image38.png image39.png media3.mp4 image40.jpeg image41.png image42.png image43.png image44.png image45.png image46.png image47.png image48.jpg image49.png image50.png image51.gif image52.png image53.png image54.png image55.png image56.png image57.png image58.png image59.png image60.png image61.png image62.png media4.mp4 image63.jpeg image64.png image250.png image65.jpg image66.png image67.png image68.png image69.png image70.png image74.png image75.png image76.png image71.png image72.png image73.png media5.mp4 image77.jpeg /docProps/thumbnail.jpeg