|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/from-0-1-machine-learning/
课程评论：没有评论
课程名称：从0到1：机器学习、自然语言处理与Python - 简明扼要 课程概述： 本课程无需任何先修知识，虽然具备一些大学数学基础会有帮助，但并非强制要求。此外，熟悉Python编程将有助于运行提供的源代码。课程由具有斯坦福大学背景的前谷歌员工和拥有印度理工学院(IIT)及印度管理学院(IIM)教育背景的前Flipkart首席分析师讲授，他们在量化交易、分析和电子商务领域拥有数十年的实际经验。 本课程旨在简单明了地介绍机器学习技巧，让学习者能够立即应用。课程的特点是：简洁易懂——尽量将复杂内容简化但不失真；自信而不浮夸——基于长期实战经验讲解，同时避免无谓的复杂性。学习后，您可以迅速上手机器学习，就像驾驭一辆汽车，而无需了解细节。 课程采用丰富的视觉元素，通过动画帮助理解，并包含大量注释的源代码，方便直接实现自然语言处理和机器学习的文本摘要和分类。设计中融入了独特的元素，例如非传统的案例、重复与内省的方式、结合小测验进行主动学习，以及活跃的音乐与艺术元素，均有助于增强认知与记忆。 主要内容： - 机器学习：监督学习与无监督学习，分类、聚类、关联检测、异常检测、降维、回归等技术，包括朴素贝叶斯、K最近邻、支持向量机、人工神经网络等。 - Python自然语言处理：语料库、停用词、句子和词语解析、自动摘要、情感分析、TF-IDF文档距离等。 - 情感分析：实用性、解决方法（基于规则的与基于机器学习的）、特征提取、情感词典、Twitter API，通过Python分析推文中的情感。 - 缓解过拟合：决策树与学习、过拟合的影响、交叉验证、正则化及随机森林等技术。 - 推荐系统：基于内容的过滤、协同过滤与关联规则学习。 - 深度学习入门：应用多层感知器解决MNIST数字识别问题。 注意事项：课程中的代码示例使用Python 2.7，并提供了丰富注释的源代码，尽量兼容Python 2和Python 3。
Prerequisites: No prerequisites, knowledge of some undergraduate level mathematics would help but is not mandatory. Working knowledge of Python would be helpful if you want to run the source code that is provided. Taught by a Stanford-educated, ex-Googler and an IIT, IIM - educated ex-Flipkart lead analyst. This team has decades of practical experience in quant trading, analytics and e-commerce. This course is a down-to-earth, shy but confident take on machine learning techniques that you can put to work today Let's parse that. The course is down-to-earth: it makes everything as simple as possible - but not simpler The course is shy but confident: It is authoritative, drawn from decades of practical experience -but shies away from needlessly complicating stuff. You can put ML to work today: If Machine Learning is a car, this car will have you driving today. It won't tell you what the carburetor is. The course is very visual: most of the techniques are explained with the help of animations to help you understand better. This course is practical as well: There are hundreds of lines of source code with comments that can be used directly to implement natural language processing and machine learning for text summarization, text classification in Python. The course is also quirky. The examples are irreverent. Lots of little touches: repetition, zooming out so we remember the big picture, active learning with plenty of quizzes. There's also a peppy soundtrack, and art - all shown by studies to improve cognition and recall. What's Covered: Machine Learning: Supervised/Unsupervised learning, Classification, Clustering, Association Detection, Anomaly Detection, Dimensionality Reduction, Regression. Naive Bayes, K-nearest neighbours, Support Vector Machines, Artificial Neural Networks, K-means, Hierarchical clustering, Principal Components Analysis, Linear regression, Logistics regression, Random variables, Bayes theorem, Bias-variance tradeoff Natural Language Processing with Python: Corpora, stopwords, sentence and word parsing, auto-summarization, sentiment analysis (as a special case of classification), TF-IDF, Document Distance, Text summarization, Text classification with Naive Bayes and K-Nearest Neighbours and Clustering with K-Means Sentiment Analysis: Why it's useful, Approaches to solving - Rule-Based , ML-Based , Training , Feature Extraction, Sentiment Lexicons, Regular Expressions, Twitter API, Sentiment Analysis of Tweets with Python Mitigating Overfitting with Ensemble Learning: Decision trees and decision tree learning, Overfitting in decision trees, Techniques to mitigate overfitting (cross validation, regularization), Ensemble learning and Random forests Recommendations: Content based filtering, Collaborative filtering and Association Rules learning Get started with Deep learning: Apply Multi-layer perceptrons to the MNIST Digit recognition problem A Note on Python: The code-alongs in this class all use Python 2.7. Source code (with copious amounts of comments) is attached as a resource with all the code-alongs. The source code has been provided for both Python 2 and Python 3 wherever possible.
