Machine Learning
August 11, 2021 2021-08-17 6:41Machine Learning
Machine Learning
Machine Learning is a key to develop intelligent systems and analyze data in science and engineering. … This course provides an introduction to the fundamental methods at the core of modern machine learning. It covers theoretical foundations as well as essential algorithms for supervised and unsupervised learning.
Every Saturday 11:00 – 13:00 (EST) starting 13 November
-
Intro to Machine Learning
Apply a machine-learning algorithm to solve a linear regression problem
- Review machine learning fundamental concepts and principles
- Recognize the general approach to finding the minima using gradient descent
- Explain the difference between the batch gradient descent and the stochastic gradient descent
- Review the concepts of data types and data structures as it pertains to Machine Learning
- Prepare a dataset so a machine learning algorithm can be applied to the data
-
Regression Models
Use the most common performance metrics used in machine learning for regression problems
- Review how to explore and manipulate data in Python
- Describe how to handle missing values in a dataset
- Recognize how regression works and different types of regression
- Describe how to create dummy variables in Python for regression problems
- Recognize the importance of measuring the model’s performance and be aware of different types of metrics available for regression problems
- Interpret the results of a linear regression model
-
Classification Models
common ML performance metrics for classification problems including Logistic Regression.
- Review the Sigmoid Function
- Identify Logistic Regression and the difference with Linear Regression
- Recognize how classification works and different types of classification
- Explain how the confusion matrix works
- Calculate the values for precision, recall and F1-score
- Build the ROC and AUC curves to measure the performance of your model
- Interpret the results of a logistic regression model
- Describe the Maximum Likelihood Estimation
- Explain how the Logistic Regression model works and its use to make prediction
- Describe the Pros and Cons of using Logistic Regression and use the model in a case study
-
Support Vector Machines
Classify linear and non-linear data using the Support Vector Machine algorithm
- Recognize general aspects of Support Vector Machine models
- Describe the general approach that a Support Vector Machine algorithm uses: hyperplanes and maximum margins
- Identify issues with the basic approach of SVM
- Recognize the approach of using Kernels in SVM to deal with non-linear classifiers.
- Use Kernels in SVM to deal with non-linear classifiers.
- Use Hyperparameter tuning to optimize the Kernels in an SVM model
- Contrast the advantages and disadvantages of SVM
-
Naive Bayes and Text Data
Build classification models using the Naive Bayes algorithm
- Recognize the Naive Bayes algorithm for making predictions in classification problems
- Use important probability concepts: conditional probability and Bayes’ Rule
- Describe the Naive-Bayes algorithm
- Use Naive-Bayes to classify new items according to characteristics of others
- Explain the Maximum a Posteriori Classification Rule
- Distinguish between text data and numeric data and explain different ways to vectorize text data into numeric data
- Use Naive-Bayes to classify texts
- Use Laplace Smoothing
- Explain Naive Bayes advantages, disadvantages and applications
-
K-nearest neighbors
Create KNN classification models
- Recognize how hybrid models work and why they can be used for either regression or classification problems
- Identify the main features of KNN
- Describe distance metrics that can be applied to different models such as KNN
- Find the best value for K, the number of neighbours in KNN
- Visualize in Jupyter Notebook how KNN performs with different data sets
- Scale your data through standardization and normalization
- Recognize the relevance of interpretability according to the problem on hand
- Describe the importance of selecting only the relevant features for your model
- Handle data according to its data type or missing values
- Recognize the Pros and Cons of KNN and run a model in a case study
-
Decision Trees
Build decision trees for classification problems.
- Recognize generalities of decision trees for machine learning
- Describe homogeneity as a measure of the importance of an attribute to help define the outcome value
- Use the Gini Index to establish the attribute that results in the best increase in homogeneity
- Use Entropy to establish the attribute that results in the best increase in homogeneity
- Avoid overfitting decision trees by using pre-pruning and post-pruning
- Recognize the advantages and disadvantages of decision trees
- Construct Decision Trees in Python
-
Ensemble Models
Apply effective machine learning models by combining models into ensembles
- Describe the logic behind ensemble methods in ML
- Apply ensembles techniques such as bagging and boosting
- Summarize how random forests work
- Utilize boosting techniques such as Adaboost and XGBoost
- Apply ensemble models for ML and review their accuracy
- Contrast the advantages and disadvantages of ensemble methods
- Discuss the concept of white-box vs black-box models
-
Clustering Techniques
Use of data sets without labelled output variables by implementing unsupervised machine learning techniques (clustering)
-
Dimensionality Reduction
Reduce the number of features in a data set to those that are best for prediction
Related Courses
$200.00
SQL
$200.00
Javascript Coding
$200.00
Security
$200.00
Data Science
$200.00