This project aims to predict machine failures using machine learning techniques. The model, developed using Python in a Jupyter Notebook environment, utilizes a combination of Keras, TensorFlow, and scikit-learn libraries for data preprocessing, model construction, training, and evaluation.
- pandas
- scikit-learn
- numpy
- TensorFlow
- Keras
The dataset used in this project is stored in a CSV file named "machine failure.csv". It consists of various features such as UDI, Product ID, Type, Air temperature, Process temperature, Rotational speed, Torque, Tool wear, and multiple failure categories.
-
Clone the Repository: Clone this repository to your local machine.
-
Install Dependencies: Ensure you have all the required dependencies installed. You can install them using pip:
pip install pandas scikit-learn numpy tensorflow keras -
Open Jupyter Notebook: Open the Jupyter Notebook file
ProjectCode.ipynbin your Jupyter Notebook environment. -
Run the Notebook: Execute each cell in the notebook sequentially to load the dataset, preprocess the data, build the machine failure prediction model, train the model, and evaluate its performance.
-
Evaluate Results: After running all the cells, review the output to observe the model's accuracy on the validation and test sets.
The neural network model architecture consists of:
- Input layer with 11 neurons (input features)
- Two hidden layers with 7 and 3 neurons, respectively, using ReLU activation function
- Output layer with 1 neuron using Sigmoid activation function for binary classification
- Validation Set Accuracy: 99.9%
- Test Set Accuracy: 99.85%
- Experiment with different neural network architectures and hyperparameters to potentially improve model performance.
- Explore additional feature engineering techniques to enhance predictive capabilities.
This project is licensed under the MIT License.