AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Machine Learning Algorithms for Regression Tasks
A lot of machine learning alcarithms and methods have been set up for classification tasks. So we actually use the neral network's output, or its confidence, in this classification task to try to assess when the network is more confident. Now, if you look over all the testing data, our accuracy is actually not great. I mean, it's something like 60 % cracy. However, our goal, if you recall, is not to try and make predictions all the time. Our goal is to try to find when we have predictability. In the top, say, 20 % most confident predictions, the network actually, as accuracy is up closer to, you know, 80