Machine Learning, the names itself says the meaning of the words. Machine learning is not a technology, it is an implimentation of technology. You will be shocked if you know when was these concepts were introduced, around 1950s and 1960s. It took many decades for these concepts to come into light. Lets strat Learning, sorry, teachiing machine to learn :p Machine learning is mainly 3 types.
- Supervised Leaerning 2.Unsupervised Learning 3. Reinforcement Learning
Supervised Learning: As we all understand what supervisor means, a person observing the task done by others and correcting/making them to understand their mistakes. Similarly, here also we human act as supervisor to the machine while it is learning and letting the machine know, hwo much mistake/error it is giving in output. Some of the supervised machine learning algorithms are linear regression, Logistic regression, Decision tree, Random Forest, KNN, Support Vecto Machine, Naive Bayes. We used labelled data to train the model. It is like everytime the machine predicts the output, we ll tell the machine this is not accurate prediction, n compute the error as we already know the output of that particular training data. We ll write algorithm, in such a way that as the machine intpu with training data, it should improve the accuracy by decreasing the error. We call this error as cost/Loss etc
Unsupervised Learning: In this type of learning process, human intervention is not required for the machine learn.If you input the trainig data, the machine will learn by that data. Examples of unsupervised learning: K means, hierarchial clustering
Reinforcement Learning: In this type of learning, machine will be subjected to an environment. The machine, here we call it agent, will automaticall determine the ideal behavior within a specific content to maximize performance
In this repository, I ll be uploading the python code without scikit learn, where actual logic behind the algorithm is implemented
Happy Learning :)