Machine Learning Fundamentals

Let us learn about Unsupervised Learning: Real-Life Applications to demonstrate the uniformity of the scikit-learn API, as well as to explain the stepss taken to solve such a problem. Let us then look at the Lesson Map which introduces us to the different topics covered in the lesson along with the lesson objectives.

Let us study about clustering, its types, basic tools and its functions.

Let us now explore a wholesale customer’s dataset, understanding the dataset, and its case study features.

Let us get introduced to data visualization, loading the dataset using Pandas, learning about the important visualization tools, and plot a histogram of one feature from the Noisy Circles dataset.

Let us now look at k-means algorithm, understanding the working of the algorithm, learning about the initialization methods, changing the number of clusters, and importing and training the k-means algorithm over a dataset.

Let us learn about Mean-Shift Algorithm, understanding the working of the algorithm, and finally import and export the Mean-Shift algorithm over a dataset.

Let us learn about DBSCAN Algorithms, understanding the working of the algorithm, its parameters, and finally import and export the DBSCAN algorithm over a dataset.

Let us now learn about evaluating the performance of clusters using available metrics in Scikit-Learn, and with the help of the Silhouette Coefficient Score and Calinski–Harabasz Index.

This video summarizes your learning of this lesson.

Let us explore the main steps for working on a supervised machine learning problem. Let us then look at the Lesson Map which introduces us to the different topics covered in the lesson along with the lesson objectives.

Let us begin the lesson with model validation and testing where we learn about data partition and its subsets of training, validation, and testing set, and its uses. Let us then learn about the split ratio and perform data partition on a simple dataset. Further, we learn about Cross Validation procedure to partition data and use it to partition the Train Set into a training and a validation set.

Let us now look at the evaluation metrics for classification tasks which includes the confusion matrix tables, its values and their explanation. Let us then find out the way in which accuracy level of confusion matrix, precision metric, and recall metric is calculated. We also look at the evaluation metrics for regression tasks and finally, calculate different evaluation metrics over a classification task.

We finally look at error analysis, Bayes error, and error analysis methodology. Then, let us understand the conditions affecting the model and finally calculating the error rate over different sets of data.

This video summarizes your learning of this lesson.

Let us cover the key steps involved in working with a supervised learning data problem. Let us then look at the Lesson Map which introduces us to the different topics covered in the lesson along with the lesson objectives.

Let us learn to explore the dataset, steps for downloading the dataset, understanding the dataset and the missing values. Further, we learn about the features and steps to preprocess the dataset.

Let us study about Naïve Bayes Algorithm, the mathematical process behind the dataset and the algorithm used, and applying it to the dataset.

Let us now learn about Decision Tree Algorithm, the process behind it, and applying it to the dataset.

Let us dive into Support Vector Machine Algorithm, the process behind the algorithm, and the rule to follow to choose the right hyperplane.

Let us now calculate the evaluation metrics for the three models using the accuracy, precision, and recall metrics in order to compare them.

This video summarizes your learning of this lesson.

Let us now focus on introducing ANNs, their different types, and the advantages and disadvantages that they present. Let us then look at the Lesson Map which introduces us to the different topics covered in the lesson along with the lesson objectives.

Let us get introduced to Artificial Neural Networks (ANN), the process behind its working, and an actual methodology to train an ANN. Let us then learn about hyperparameters and the commonly used hyperparameters followed by understanding the applications and limitations of ANNs.

Let us now work on applying an ANN using Scikit-Learn's Multilayer Perceptron and apply it on the Classifier class.

Finally, let us learn about performance analysis using error analysis, fine-tuning the hyperparameters, and model comparison.

This video summarizes your learning of this lesson.

Let us now build a complete machine learning program in this lesson. Let us then look at the Lesson Map which introduces us to the different topics covered in the lesson along with the lesson objectives.

Let us begin with defining a machine learning program and defining its stages of preparation, creation, and interaction. Next, we try to understand the dataset at work.

Next, let us understand and work on two important actions of saving and loading a dataset so that it can be reused at any moment through different means.

Finally, let us outline the steps involved in interacting with a trained model, and creating a class and channel to interact with trained model.

This video summarizes your learning of this lesson.