Optimized Convolutional Neural Network Model for Software Effort Estimation

Optimized Convolutional Neural Network (CNN) designed as a useful tool for approximating and forecasting software effort. Convolutional Neural Networks, a subfield of machine learning, are utilized in estimate due to its propensity for quick learning and higher-quality, more accurate output. Using the effort estimation while also determining the most straightforward Convolutional neural network architecture for optimal learning. Additionally, a range of machines, such as convolutional neural networks, decision trees, random forests, principal component regression and support vector machines, are utilized to determine which model performs best in terms of software effort estimation. This experiment made use of the JM1 dataset. The metrics under discussion are evaluated using the following measures: Mean Absolute Error (MAE), Mean Squared Error (MSE), R-squared and Pred (25). Convolutional Neural Networks are the best models in this technique, as demonstrated by the results when compared to other models. Estimating software effort is used to project planning and resource allocation and project management as a whole estimation. Conventional models used for estimate techniques have limited accuracy and flexibility since they frequently rely on old data and judgment.