Whenever we think about model prediction, the first thing pops up is always prediction errors(bisa and variance). Understanding of these errors would help us avoid the mistake of overfitting and underfitting.
[Read More]
All you need know about those models
Key components in statistics
P_value:An important metric for feature selection since it shows the correlation between target and features(variables)Example Here, I want to address the importance of feature selection: It enables the machine learning algorithms to train faster It improves the accuracy of a model if the right sublet is chosen It reduces overfitting...
[Read More]
Infamous Logistic Regression
What exactly is logistic regression. For binary classification, logistic regression computes the posterior probability of class C1 as a logistic sigmoid acting on a linear function of the feature vector x: p(C1|x)=y(x)=σ(wTx).
[Read More]
Gaussian Classifiers and Naive Bayes
Two Generative Models
By the end of this article, you would understand how them work. A generative model is actually estimating p(x|y) to then deduce p(y|x).In this case,if we learn the class-conditonal density p(x|Ck) through training, we are able to predict the label for the datapoint x p(Ck|x) = p(x|Ck)p(Ck)/p(x) Confused by all...
[Read More]
All you need know about reinforcement learning algorithms
In reinforcement learning, the output is an action or sequence of actions which is decided by reward or penalty. The goalis to find the best policy to maxmize the expected sum of the future reward. We also use a discount factor because we dont have to take future reward too...
[Read More]