Day 9 : Progressed with Course 2 of Andrew Ng " Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization " and learnt about and googled up read theory part from the book on the following keywords
- layer
- hidden units
- learning rates
- activation functions
- hyperparameters
- bias
- variance
- overfitting
- underfitting
- high variance
- high bias
- low bias
- low variance
- Regularization
- why regularization reduce overfitting
For the practical hands on part didnt code today instead signed up for IBM cloud and tried small hands on image label classification using the pretrained model on IBM watson cloud instance , will be be trying the both the cloud service API , deploying custom model in production on cloud for a building a better understanding of the data pipeline [also try including that in my Project] , for the NLP manual part will practise some simple python programs for better understanding and creating own scripts for filtering the data , also reading and watching tutorial absorbs the data but should make some hard notes to refer back and add on to when building on those subjects for a better mind map. With the IBM cloud hands on completed one of the courses on IBM applied course , the hands on programming part for today wasnt sufficient so need to plan schedule accordingly
Certificate Link
No comments:
Post a Comment