|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/crash-course-on-machine-learning-and-intro-to-deep-learning/
课程评论：没有评论
课程名称：掌握人工智能 - 机器学习与深度学习简介 课程概述：本课程是一门深入且速成的课程，旨在培养学员成为机器学习专家。课程设计旨在解决现实生活中的问题，能够为当前的项目和所在组织带来实质性帮助。无论是学生、教授还是机器学习顾问，都能在这门课程中找到乐趣，课程内容简单易懂且与时俱进。完成课程后，学员将成为可就业的机器学习工程师和数据科学家。该课程由一位热情且经验丰富的教授授课，经过多次课堂和项目测试，确保了课程的实用性和有效性。学员将在课程中进行多个项目的实践。 课程内容分为七个模块，采用丰富的案例和示例，通过在Jupyter Notebook上执行代码进行展示。课程的第一模块侧重于数据可视化，强调在进行机器学习之前可视化数据的重要性。接下来，课程将深入介绍回归分析，包括简单线性回归、Polynomial回归和多重线性回归。在回归分析后，课程将详细讨论分类的监督学习算法，包括逻辑回归、朴素贝叶斯分类器、支持向量机、K近邻、决策树、集成学习、分类与回归树、随机森林以及提升算法（如AdaBoost、梯度提升等）。 然后，课程将转向无监督学习，讨论无标签数据的聚类问题，包括层次聚类、K均值聚类、K Medoids聚类以及聚合聚类。此外，课程还将讲解在机器学习领域中成功所需的策略，如偏差-方差权衡和维度诅咒。学员还将学习主成分分析（PCA）和线性判别分析（LDA），以应对维度诅咒的问题。 课程的最后部分将简要介绍深度学习，包括人工神经网络（ANN）及其反向传播算法，帮助学员了解前馈网络的权重估计。在结束课程之前，学员将进行两个案例研究，一个是二分类，另一个是多分类，使用ANN来让学员更好地体验深度学习的应用。
This is a crash course, but an in-depth course, which will develop you as a Machine learning specialist. Designed with solutions to real life life problems, this will be a boon for your ongoing projects and the organization you work for. Students, Professors and machine learning consultants will find the course interesting, hassle free and up-to-date. Surely, the students will be employable Machine Learning Engineers and data scientists. Given by an enthusiastic and expert professor after testing it in classrooms and projects several times. The students can carry out a number of projects using this course. This exemplary, engaging, enlightening and enjoyable course is organized as seven interesting modules, with abundant worked examples in the form of code executed on Jupyter Notebook. It is important that data is visualized before attempting to carryout machine learning and hence we start the course with a module on data visualization. This is followed by a full blown and enjoyable exposure to Regression covering simple linear regression, polynomial regression, multiple linear regression. Regression is followed by extensive discussions on another important supervised learning algorithms on Classification. We carry out modeling using classification strategies such as logistic regression, Naive Bayes classifier, support vector machine, K nearest neighbor, Decision trees, ensemble learning, classification and regression trees, random forest and boosting - ada boost, gradient boosting. From supervised learning we move on to discuss about unsupervised learning - clustering for unlabelled data. We study the hierarchical, k means, k medoids and Agglomerative Clustering. It is not enough to know the algorithms, but also strategies such as bias variance trade off and curse of dimensionality to be successful in this challenging field of current and futuristic importance. We also carry out Principal Component analysis and Linear discriminant analysis to deal with curse of dimensionality.The last section leads the reader to deep learning through a lucid introduction to Artificial Neural network (ANN) and back propagation algorithm for estimating weights of feed forward network. Before we close, we take up 2 case studies- one on binary classification and another on multi-class classification using ANN, to give a feel of deep learning.
