machine learning

machine learning

Porter Stemmer algorithm

Stemming is the process of reducing a word to its stem that affixes to suffixes and prefixes or to the roots of words lemma. We cover the algorithmic steps in Porter Stemmer algorithm, a native implementation in Python, implementation using Porter Stemmer algorithm from NLTK library and conclusion.

machine learning

Bias in Machine learning

Bias is an constant parameter in the Neural Network which is used in adjusting the output. Therefore Bias is a additional parameter which helps the model so that it can perfectly fit for the given data. It is also known as bias nodes, bias neurons, or bias units

machine learning

Implementing CNN in Python with Tensorflow for MNIST digit recognition

In this article, we will develop and train a convolutional neural network (CNN) in Python using TensorFlow for digit recognifition with MNIST as our dataset. We will give an overview of the MNIST dataset and the model architecture we will work on before diving into the code.

machine learning

Logistic Regression in Python with TensorFlow

We will walk you though the difference between Linear and Logistic Regression and then, take a deep look into implementing Logistic Regression in Python using TensorFlow. We used the Iris dataset and have trained and plotted the loss function and the training and test accuracy across epochs

machine learning

Training, saving and loading Artificial Neural Networks in Keras

We demonstrate how to code a Artificial neural network model and train and save it in JSON or H5 format which can be loaded later for any inference task. We use Keras/ TensorFlow to demonstrate this transfer learning and used Pima Indian Diabetes dataset in CSV format

machine learning

Random Decision Forest

Random forests or random decision forests are an ensemble learning method for classification, regression and other tasks that operates by constructing a multitude of decision trees at training time. There are three important hyperparameters namely n_estimators, random_state and max_features

machine learning

Generative Model

A Generative Model is a way of learning any kind of data distribution. Generative modeling algorithms process the training data and make reductions in the data. The main aim is to learn the true data distribution of the training set so that the new data points are generated with some variations.

machine learning

Discriminative Model

Discriminative models, also referred to as conditional models, are a class of models used in statistical classification, especially in supervised machine learning. Discriminative modelling studies the P(y|x) i.e, it predicts probability of y(target) when given x(training samples).

machine learning

XGBoost

XGBoost is short for extreme gradient boosting. It is an optimized distributed gradient boosting library designed to be highly efficient, flexible and portable. It implements machine learning algorithms under the Gradient Boosting framework. It provides a parallel tree boosting known as GBDT, GBM

machine learning

Independent Component Analysis (ICA)

Independent component analysis (ICA) is a statistical and computational technique for revealing hidden factors that underlie sets of random variables, measurements, or signals and is a special case of blind source separation. A common application is to listen to one person's speech in a noisy room

machine learning

Linear Regression in Python with TensorFlow

In this guide, we will implement Linear Regression in Python with TensorFlow. Linear Regression is a simple yet effective prediction that models any data to predict an output based on the assumption that it is modeled by a linear relationship.

machine learning

Principal Component Regression (PCR)

Principal Component Regression (PCR) is an algorithm for reducing the multi-collinearity of a dataset. PCR is basically using PCA, and then performing Linear Regression on these new PCs. The key idea of how PCR aims to do this, is to use PCA on the dataset before regression.

machine learning

Ridge Regression

Ridge regression is an efficient regression technique that is used when we have multicollinearity or when the number of predictor variables in a set exceed the number of observations. It uses L2 regularization and solves the problem of overfitting. Concepts of overfitting and regularization is basis

machine learning

Model Evaluation: a crucial step in solving a machine learning problem

Models like Googlenet is used across various problems and MobileNet are designed for computational limited resources. It is a challenge to find the best technique or model for a given problem. We evaluate a model based on Test Harness, Performance Measure, Cross validation and Testing Algorithms.

machine learning

Summary of Regression Techniques

Regression is a technique based on statistics to model the relationship between a set of variables to make predictions on unseen data. We explored are Linear, Logistic, Polynomial, Ridge, Lasso, Elastic Net, Stepwise regression.

machine learning

Data augmentation Techniques

Data augmentation is the technique of increasing the size of data used for training a model. Some of position augmentation includes scaling, cropping, flipping, padding, rotation, translation, affine transformation. Color augmentation includes brightness, contrast, saturation and hue.

algorithm

Expectation Maximization Clustering Algorithm

Expectation Maximization Clustering algorithm is much more robust than K-Means, as it uses two parameters, Mean and Standard Deviation to define a particular cluster. This simple addition of calculating the Standard Deviation, helps the EM algorithm do well in a lot of fail cases of K-Means

algorithm

Mean Shift Clustering Algorithm

Mean Shift clustering is an unsupervised clustering algorithm that groups data directly without being trained on labelled data. It is hierarchical in nature. It starts off with a kernel, which is basically a circular sliding window. The bandwidth the radius of this sliding window is pre-decided

machine learning

Autoencoder

An autoencoder is a neural network that learns data representations in an unsupervised manner. Its structure consists of Encoder, which learn the compact representation of input data, and Decoder, which decompresses it to reconstruct the input data.

machine learning

Neural Style Transfer using CNN

We demonstrate the easiest technique of Neural Style or Art Transfer using Convolutional Neural Networks (CNN). We use VGG19 as our base model and compute the content and style loss, extract features, compute the gram matrix, compute the two weights and generate the image with the other style

machine learning

Fully Connected Layer: The brute force layer of a Machine Learning model

Fully Connected layers in a neural networks are those layers where all the inputs from one layer are connected to every activation unit of the next layer. In most popular machine learning models, the last few layers are full connected layers which compiles the data extracted by previous layers

machine learning

Convolution Layer: The layer that takes over 70% of time in a Machine Learning model

Convolutional Layer is the most important layer in a Machine Learning model where the important features from the input are extracted and where most of the computational time (>=70% of the total inference time) is spent. Concepts involved are kernel size, padding, feature map and strides

clustering algorithm

K+ Means Clustering algorithm

K+ Means algorithm is a clustering algorithm and an improvement to K means clustering algorithm and solves the problem of choosing K (number of clusters). It is great at detecting outliers and forming new clusters. The complexity is O(t*(k^2)*n) which is slightly more than K means algorithm

machine learning

Types of Data Formats in Machine Learning

Each data format represents how the input data is represented in memory. This is important as each machine learning application performs well for a particular data format and worse for others. Various data formats are NHWC, NCHW, NCDHW and NDHWC

machine learning

Types of Loss Functions in Machine Learning

The various types of loss functions are mean_squared_error, mean_absolute_error, mean_absolute_percentage_error, mean_squared_logarithmic_error, squared_hinge, hinge, categorical_hinge, logcosh, categorical_crossentropy, sparse categorical / binary crossentropy, kullback_leibler_divergence and other