Exam PA – Predictive Analytics
5个单元 · 22个章节
免费的Exam PA – Predictive Analytics学习笔记，专为SOA (Society of Actuaries)学生准备。下列每个章节都涵盖一个重点主题，附有例题与练习提示，可在 thinka 应用中延伸练习。
Predictive Analytics Problem Definition
Descriptive, predictive, and prescriptive analytics
Characteristics of predictive modeling problems
Bias, variance, model complexity, and the bias-variance trade-off
Translating a vague question into an analyzable one
Defining the problem: data, technology, business impact, and implementation
Identifying additional information and next steps
Data Exploration and Visualization
Structured and unstructured data types
Types of variables and predictive modeling terminology
Data design: time frame, sampling, and granularity
Key principles of constructing graphs
Univariate data exploration
Bivariate data exploration
Data Transformations and Unsupervised Learning Techniques
Creating features from existing data
Principal components analysis for data transformation
K-means and hierarchical clustering for data transformation
Generalized Linear Models
Selecting and validating a GLM for a business problem
Offsets and weights
Interpreting model coefficients and interaction terms
Hyperparameters for regularized regression
Tree-Based Models
Constructing, pruning, and validating regression and classification trees
Bagging and random forests
Boosting
看完笔记了？用AI练习题测试自己
立即练习此课题