Exam PA – Predictive Analytics

5 个单元 · 22 个章节

免费 Exam PA – Predictive Analytics 学习笔记，适合 SOA (Society of Actuaries) 学生。每个章节都覆盖一个重点主题，并附例题与可延伸到 thinka app 的练习提示。

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

把笔记变成真正的练习

不要只停留在阅读。立即用无限 AI 题目练习，拿到即时批改与清楚反馈。

立即开始练习查看收费

看完笔记了？现在就用 AI 题目测一测自己。

马上练这一题