Dear scholars,
Do you have any suggestions for a concise and useful book with examples covering continuous numerical value predictions?
Suppose that, we need to use regression techniques for such predictions. Due to non-linearity, and complexity of features and output; linear regression or decision trees or K-nearest neighbors are not good fits. So, neural network is the choice. I this regard, do you know any concise books?
Thanks!
You can start from these two websites ( https://towardsdatascience.com/simply-deep-learning-an-effortless-introduction-45591a1c4abb, https://towardsdatascience.com/everything-you-need-to-know-about-neural-networks-and-backpropagation-machine-learning-made-easy-e5285bc2be3a). This book ( https://www.springer.com/gp/book/9789811358494 ) is also a useful resource.
I recommend you before going to learn another models, try ensemble-learning techniques!
Knn and tree models are prone to overfitting!
In ensemble-learning they are classified as strong lerners.
Besides, linear models might be a weak learners.
Why don't you use ensemble-learning techniques to boost your models.
Bagging is good for models like tree and knn
And Adaboost is good for weak learners!
These techniques are so powerful and may be helpful for you.
Try them!
- Badges
- Science topic
- Similar topics
- Books