文件列表:

Code/
Dataset/
Result/

# Neuronal and Evolutionary Computation (MIA-MEISISI) Practical Exercise Students Name: MD ARIF ULLAH BHUIYAN & JULIO CSAR SIGUEAS PACHECO GitHub Link: [A1 Prediction with Back-Propagation and Linear Regression](https://github.com/arifbhuiyan82/A1-Prediction-with-Back-Propagation-and-Linear-Regression) Objective Perform data classification with the following algorithms: 1. Neural Network with Back-Propagation (BP), custom Implementation. 2. Neural Network with Back-Propagation (BP-F), using free software. 3. Multiple Linear Regression (MLR-F), using free software. Algorithms Overview # Neural Network with Back-Propagation (BP) The Neural Network with Back-Propagation, often referred to as Back-Propagation (BP), is a supervised learning algorithm used for training Multi-Layer Perceptrons (MLPs). It involves both forward and backward passes, where inputs are passed through the network to generate an output, and errors are propagated back through the network to update the weights. It is widely used for pattern recognition, classification, and regression tasks. # Neural Network with Back-Propagation with Momentum (BP-F) This variant of the standard BP algorithm includes a momentum term in the weight update rule. The momentum term accelerates the learning process by considering the previous change in weight along with the current gradient. It helps prevent the network from getting stuck in local minima and can result in faster convergence. # Multiple Linear Regression with Feature Selection (MLR-F) Multiple Linear Regression (MLR) models the relationship between a dependent variable and multiple independent variables by fitting a linear equation to the data. The 'F' typically indicates the involvement of feature selection, which simplifies the model by selecting relevant features and removing irrelevant ones. Datasets Three datasets are used in this exercise: 1. A1-synthetic.txt: This dataset contains 1001 entries across 10 columns, with v1 to v9 as independent variables and column z as the target variable. 2. A1-turbine.txt: This dataset consists of 452 entries across 5 columns, with features like height_over_sea_level, fall, net_fall, flow, and power_of_hydroelectrical_turbine. 3. winequality-white.csv: This dataset contains 750 rows and 12 columns, representing characteristics of white wines, including features like acidity, sulfur dioxide levels, alcohol, and quality ratings. Data Preprocessing Data normalization is performed using the MinMaxScaler to bring all data values to a consistent scale within the range [0, 1]. This process ensures stable training and avoids dominance of features based on magnitude. Implementation of BP-F The implementation of BP-F includes: - Initialization of random weights and biases. - Forward pass for prediction. - Loss calculation using mean squared error. - Backward pass to update weights and biases. - Training the neural network with multiple epochs. - Making predictions using the trained model. Plots The code includes two plots for visualization: 1. Training Losses Plot: This plot displays the training errors over epochs, helping visualize the learning progress. 2. Predictions vs. Actual Targets Plot: This scatter plot shows the predictions against actual target values in the test set, allowing you to assess the model's performance. Feel free to explore the code and datasets in this repository to learn more about these algorithms and their applications.

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：