Introduction to Machine Learning

Introduction to Machine Learning

what is machine learning

1. Machine learning is a multi-disciplinary subject, including but not limited to computers, mathematics, and statistics. Specializing in how computers or other software and hardware devices simulate or realize human learning behaviors to acquire new knowledge or skills, reorganize existing knowledge structures, and continuously improve their performance.
2. Machine learning is the core content of artificial intelligence research. There are numerous applications of machine learning in all branches of artificial intelligence. For example: "expert system", "automatic reasoning", "natural language", "pattern recognition" (pattern: recognition of common patterns), "computer vision", "intelligent robot" and other fields.
3. Applying machine learning technology to products brings users the shocking experience of "machines possess human intelligence".
4. Human costs are rising, and machine learning can reduce business costs.
5. The second machine revolution - the dominance of machines with human intelligence as the core value
6. Machine learning is widely used in data mining, and its machine connotation is almost universal, which can be regarded as different sides of the same mountain.

Fields with many machine learning applications

1. Data Analysis and Data Mining: Implementing a set of tools, methods, or procedures using machine learning to extract valuable knowledge, rules, and patterns from real-world content data. And apply the results to the front-end system to assist the business, such as: recommendation, decision-making system, accurate identification, participation in services, etc.
2. Image and voice recognition: voice input, OCR, handwriting input, communication monitoring, license plate recognition, fingerprint recognition, iris recognition, face recognition, etc.
3. Smart machines, robots: production line robots, man-machine dialogue, computer games.
4. Text mining: spam filtering, news scraping, comment analysis (Bayesian classifier).

software

R 、 Apply 、 Matlab 、 Python

representative algorithm

1. Regression prediction and corresponding dimensionality reduction techniques: linear regression, logistic regression, principal component analysis, factor analysis, ridge regression, LASSO
2. Classifiers: Decision Trees, Naive Bayes, Bayesian Belief Networks, Support Vector Machines, Adaboost and Random Forest to improve classifier accuracy
3. Clustering and Outlier Discrimination
4. Artificial neural networks