Visão Geral

O Curso Comprehensive Machine Learning with Python se baseia em nosso curso Comprehensive Data Science with Python e ensina aos participantes como escrever aplicativos de aprendizado de máquina em Python.

Objetivo

Após realizar este Curso Comprehensive Machine Learning with Python você será capaz de:

• Entenda o aprendizado de máquina como uma ferramenta útil para modelos preditivos
• Saiba quando recorrer ao aprendizado de máquina como ferramenta
• Implementar o pré-processamento de dados para um fluxo de trabalho de ML
• Entenda a diferença entre tarefas supervisionadas e não supervisionadas
• Implementar vários algoritmos de classificação
• Avalie o desempenho do modelo usando uma variedade de métricas
• Compare modelos em um fluxo de trabalho
• Implementar variações do algoritmo de regressão
• Compreenda as abordagens de clustering para dados
• Interpretar rótulos gerados a partir de clustering
• Transforme dados de texto não estruturados em dados estruturados
• Compreender a preparação de dados específicos de texto
• Visualize dados de frequência de fontes de texto
• Execute a modelagem de tópicos em uma coleção de documentos
• Use texto rotulado para realizar a classificação de documentos

Materiais

Conteúdo Programatico

Introduction

Review of Core Python Concepts

1. Anaconda Computing Environment
2. Importing and manipulating Data with Pandas
3. Exploratory Data Analysis with Pandas and Seaborn
4. NumPy ndarrays versus Pandas Dataframes

An Overview of Machine Learning

1. Machine Learning Theory
2. Data pre-processing
3. Missing Data
4. Dummy Coding
5. Standardization
6. Data Validation Strategies
7. Supervised Versus Unsupervised Learning

Modeling for explanation (descriptive models)

1. Understanding the linear model
2. Describing model fit
3. Adding complexity to the model
4. Explaining the relationship between model inputs and the outcome
5. Making predictions from the model

Supervised Learning: Regression

1. Linear Regression
2. Penalized Linear Regression
3. Stochastic Gradient Descent
4. Decision Tree Regressor
5. Random Forest Regression
6. Gradient Boosting Regressor
7. Scoring New Data Sets
8. Cross Validation
9. Variance-Bias Tradeoff
10. Feature Importance

Supervised Learning: Classification

1. Logistic Regression
2. LASSO
3. Support Vector Machine
4. Random Forest
5. Ensemble Methods
6. Feature Importance
7. Scoring New Data Sets
8. Cross Validation

Unsupervised Learning: Clustering

1. Preparing Data for Ingestion
2. K-Means Clustering
3. Visualizing Clusters
4. Comparison of Clustering Methods
5. Agglomerative Clustering and DBSCAN
6. Evaluating Cluster Performance with Silhouette Scores
7. Scaling
8. Mean Shift, Affinity Propagation and Birch
9. Scaling Clustering with mini-batch approaches

Clustering for Treatment Effect Heterogeneity

1. Understand average versus conditional treatment effects
2. Estimating conditional average treatment effects for a sample
3. Summarizing and Interpreting

Data Munging and Machine Learning Via H20

1. Intro to H20
2. Launching the cluster, checking status
3. Data Import, manipulation in H20
4. Fitting models in H20
5. Generalized Linear Models
6. naïve bayes
7. Random forest
8. Gradient boosting machine (GBM)
9. Ensemble model building
10. automl
11. data preparation
12. leaderboards
13. Methods for explaining modeling output

Introduction to Natural Language Processing (NLP)

1. Transforming Raw Text Data into a Corpus of Documents
2. Identifying Methods for Representing Text Data
3. Transformations of Text Data
4. Summarizing a Corpus into a TF—IDF Matrix
5. Visualizing Word Frequencies

NLP Normalization, Parts-of-speech and Topic Modeling

1. Installing And Accessing Sample Text Corpora
2. Tokenizing Text
3. Cleaning/Processing Tokens
4. Segmentation
5. Tagging And Categorizing Tokens
6. Stopwords
7. Vectorization Schemes for Representing Text
8. Parts-of-speech (POS) Tagging
9. Sentiment Analysis 
10. Topic Modeling with Latent Semantic Analysis

NLP and Machine Learning

1. Unsupervised Machine Learning and Text Data
2. Topic Modeling via Clustering
3. Supervised Machine Learning Applications in NLP