Skip to content

Latest commit

 

History

History
20 lines (15 loc) · 909 Bytes

File metadata and controls

20 lines (15 loc) · 909 Bytes

CPU Performance Prediction using MLP

This project aims to create a model which predicts the relative performance of a CPU based on some characteristics. The input-output mapping is nonlinear, thus a Multi-Layer Perceptron (MLP) is created using TensorFlow and Keras ML frameworks. The input layer takes a 6-feature input vector. We designed a 2-hidden layer architecture of 64 neurons each. Regarding the output layer, we selected a linear activation function due to the nature of the problem (unbounded output - regression problem).

Libraries/Frameworks used

  • Pandas
  • NumPy
  • Seaborn
  • Matplotlib
  • scikit-learn
  • TensorFlow
  • Keras

Results

image

image