DL_Lab_1_Wiki
The Deep Learning Lab - 1 task is to implement Logistic Regression on the dataset using TensorFlow.
The dataset chosen to perform the logistic regression identifies the risk that is associated with giving birth to the child who is low in weight (less than 2500 kilograms). The logistic regression algorithm is chosen as we should predict whether the child is a low weight baby or not which is a Boolean situation.
The Variables which are used for prediction are:
- Age of mother
- Weight during the last menstrual period
- Race
- Smoke
- History of the pre-mature labor
- Hyper Tension
- Uterine Irritability
- Birth weight in grams
The Screenshots of the code are:
The Screenshot of the result is:
The graphs that evaluate the performance of the model are:
Cross-Entropy Loss per Generation:
Training and Test Set Accuracy:
By analyzing these two graphs we can say that the model performs better as we can observe there is a gradual decrease in loss over the generations and also the improvement in the accuracy of the model.
The Link to the Lab -1 Report is: DL_Lab_1_Report