Machine Learning Tips and Tricks

Difference between classification and regression In the above diagram the left side picture depicts classification(Disease, Healthy) and the right side picture depicts the regression. Classification: the output variable takes class labels. Regression: the output variable takes continuous values. Regression and classification are both related to prediction, where regression predicts a value from a continuous set, whereas classification predicts the 'belonging' to the class. For example, the price of a house depending on the 'size' , can be some 'numerical value' (which can be continuous): this relates to regression. Similarly, the prediction of price can be in words, viz., 'very costly', 'costly', 'affordable', 'cheap', and 'very cheap': this relates to classification. Each class may correspond to some range of values.

- Supervised learning  - Unsupervised learning - Others being Reinforcement learning - Supervised learning : In supervised learning the data is a pre-classified one ie., x->y. Here the problem is "given 'x' we need to predict 'y' " In unsupervised learning we have no idea what our results should look like ie., we know only "x" Ex: Problem statement - Basket of fruits to arrange the same type of  fruits at one place. In the above problem statement what is supervised learning? and unsupervised learning? Supervised -- Arranging same type of fruits at one place  knowing what are all the fruits Unsupervised -- Arranging same type of fruits at one place without knowing the fruits and first time we are seeing

- Database mining (process of discovering interesting patterns in large amounts of data) Ex: Web click data, medical records data, biology, engineering data - Applications can't program by hand Ex: Autonomous driving, hand written recognition, natural language processing(NLP), computer vision. - Self-customizing programs Ex: Amazon, flipkart, Netflix movie recommendations, product recommendations - Understanding human learning (mimic human brain) - Virtual personal assistants (chat bots)   - Online fraud detection - spam filters   - Medical diagnosis  - Trading strategies - supply chain management

Arthur Samuel   -  Machine learning is a field of study that gives computers the ability to learn without being explicitly programmed. Explanation : With machine learning we can write a program that will learn things on its own, so with one program for example it can learn to detect if a there is a cat in the picture. Now we don't have to tell the program if each picture has a cat in it or not, the program will learn that itself using machine learning (no explicit programming required). Tom Mitchell   -  A computer program is said to learn from experience  E with respect to some task  T and some performance measure by P, improves with experience E. Explanation:    down vote Machine learning is a term that is taken from the real world  problem of a person, and applied on something that can't actually learn - a machine. - machine learning will not (usually) change the code, but it might change it's execution path and decision based on previous data

1. Machine learning is an application of Artificial intelligence that give computers the ability to learn without being explicitly programmed. 2.  With machine learning we can write a program that will learn things on its own, so with one program for example it can learn to detect if a there is a cat in the picture. Now we don't have to tell the program if each picture has a cat in it or not, the program will learn that itself using machine learning (no explicit programming required). 3. Machine Learning algorithms can be  generalized ie., one machine learning program may serve for many domains. Eg: Object recognition - If we write a machine learning program to detect a cat in a picture, the same program may be used to detect other objects also(no explicit programming).