BLOG

Where I collect my personal thoughts and findings about my projects

__

Building a Python class to visualize the internal process of a neural network

December 29, 2019

This post develops a Python class to visualize what happens inside a feed-forward neural network when the input is the full 2D space.

__

How does a neural network internally shape the space?

December 22, 2019

This post shows what happens inside a trained feed-forward NN when the input is the full 2D space.

__

Can we visualize the flow of a multiclass neural network?

December 15, 2019

This post shows what happens inside different types of feed-forward neural network for a multi-class classification problem.

__

Can we visualize the flow of a regression neural network?

December 8, 2019

This post shows what happens inside different types of feed-forward neural network for a regression problem.

__

How to build a Python class to visualize a neural network?

December 1, 2019

This post develops a Python class to visualize what happens inside a feed-forward neural network.

__

Can we see inside a neural network?

November 24, 2019

This post shows what happens inside a trained feed-forward NN.

__

Multi-hyperparameter analysis of a neural network and computational comparison

November 17, 2019

This post looks for the best set of hyperparameters and compares the computational effort of each of the three machine-learning libraries.

__

Hyperparameter analysis for regression

November 10, 2019

This post analyzes the hyperparameter (HP) space for a regression problem in Keras.

__

Hyperparameter analysis for multi-class classification

November 3, 2019

This post analyzes the hyperparameter (HP) space for a multi-class classification problem in Keras.

__

Meta-learning neural networks over basic tasks

October 27, 2019

This post explores the *meta-learning* concept and shows how to achieve this goal with our internally-developed Python class.

__

How neural networks learn basic features with Pytorch

October 20, 2019

This post shows how to train a neural network on some basic examples with Pytorch.

__

Key notions of Pytorch

October 13, 2019

This post introduces the key components and modules of Pytorch, which are going to be applied to a neural network.

__

How neural networks learn basic features with Tensorflow

October 6, 2019

This post shows how to train a neural network on some basic examples with Tensorflow.

__

How neural networks learn basic features with Keras

September 29, 2019

This post shows how to train a neural network on some basic examples with Keras.

__

How neural networks learn basic features with Scikit-learn

September 22, 2019

This post shows how to train a neural network on some basic examples with Scikit-learn.

__

How neural networks learn basic features - Create datasets

September 15, 2019

This post shows how to create a dataset for three different basic applications, regression, binary- and multi-classification, which a fully-connected neural network needs to learn.

__

FCNN - Geometric intuition behind neural networks

September 8, 2019

This post gives some geometric insight into what occurs in a fully-connected neural network.

__

FCNN - Geometric intuition behind a neuron

September 1, 2019

This post gives some geometric insight into what occurs in a single neuron of a FCNN.

__

Fully connected neural networks - cheat sheet

August 25, 2019

This post wants to be a cheat sheet for fully-connected neural networks. It should be good either for a fresher to start and for a practitioner to quickly review.

__

Code art in Python - Spirograph pattern in a rectangle - Part 2

August 18, 2019

This post shows how to identify the polygon vertexes, post-process the intersection points, define a colour scheme, implement the global function to draw a spirograph pattern in a rectangle with Python. Some drawings will be shown at the end of the post.

__

Code art in Python - Spirograph pattern in a rectangle - Part 1

August 11, 2019

This post defines the grid definition and the parametric equations required to draw a spirograph pattern in a rectangle with Python.

__

Code art in Python - Spirograph pattern in circle - Part 4

August 4, 2019

This post defines the polygon drawing, different colour schemes, implements the global function to draw a spirograph pattern in a circle with Python. Some drawings will be shown at the end of the post.

__

Code art in Python - Spirograph pattern in circle - Part 3

July 28, 2019

This post implements a vectorized polygon vertex detection and shows how to postprocess intersection points to create a spirograph pattern in a circle with Python.

__

Code art in Python - Spirograph pattern in circle - Part 2

July 21, 2019

This post shows how to define a line through two points and the intersection point of two lines.

__

Code art in Python - Spirograph pattern in circle - Part 1

July 14, 2019

This post defines how to create a spirograph pattern in a circle with Python and shows the required visualization basics.

__

Learning Python by doing - Part 5

July 7, 2019

This post develops a Python code to identify the highest employee wages in a company and assesses each method's computational performances.

__

Learning Python by doing - Part 4

June 30, 2019

This post develops a Python code from scratch to identify the highest employee wages in a company.

__

Learning Python by doing - Part 3

June 23, 2019

This post develops a Python code from scratch to process text. How to extract initials from a name string.

__

Learning Python by doing - Part 2

June 16, 2019

This post develops a Python code from scratch to determine the travelled distance of a bike rider in a given time span, if the annual target keeps increasing.

__

Learning Python by doing - Part 1

June 9, 2019

This post shows how to develop some Python code from scratch to determine how many kilometers a bike rider will cover in a given time span.

__

Do not overlearn your data too much, learn to generalize - Part 8

June 2, 2019

This post analyses the effects of regularization on a non-linear binary classification problem.

__

Do not overlearn your data too much, learn to generalize - Part 7

May 26, 2019

This post analyses the effects of regularization on a non-linear binary classification problem.

__

Do not overlearn your data too much, learn to generalize - Part 6

May 19, 2019

This post analyses the effects of regularization on a linear binary classification problem.

__

Do not overlearn your data too much, learn to generalize - Part 5

May 12, 2019

This post analyses the effects of regularization on a linear binary classification problem.

__

Do not overlearn your data too much, learn to generalize - Part 4

May 5, 2019

This post implements the Ridge regression in Python from scratch. It is applied to a low-degree and high-degree model, compared to a non-regularized model and it is optimized on the validation set.

__

Do not overlearn your data too much, learn to generalize - Part 3

April 28, 2019

This post goes through the hyperparameter optimization for model selection, via cross-validation, and the regularization technique. The former topic is more recently referred to as **meta-learning**.

__

Do not overlearn your data too much, learn to generalize - Part 2

April 21, 2019

This post aims at visualizing the bias-variance dilemma, understanding how the model capacity relates to its performance and why it is common practice to split the dataset into training and testing, creating some learning curves that should clarify whether gathering additional data might be worthy.

__

Do not overlearn your data too much, learn to generalize - Part 1

April 14, 2019

This post defines the bias and variance concepts and apply it to both a linear and a non-linear model.

__

Learning to classify coffee from cappuccino - Part 8

April 7, 2019

This post implements a logistic regression in Scikit-learn for the MNIST dataset.

__

Learning to classify coffee from cappuccino - Part 7

March 31, 2019

This post implements a logistic regression in Scikit-learn for the digits dataset.

__

Learning to classify coffee from cappuccino - Part 6

March 24, 2019

This post implements a multinomial logistic regression with categorical inputs in Scikit-learn.

__

Learning to classify coffee from cappuccino - Part 5

March 17, 2019

This post implements a multinomial logistic regression with Scikit-learn.

__

Learning to classify coffee from cappuccino - Part 4

March 10, 2019

This post implements a multivariate logistic regression algorithm with non-linear predictors with Scikit-learn.

__

Learning to classify coffee from cappuccino - Part 3

March 3, 2019

This post implements a logistic regression algorithm in Scikit-learn and TensorFlow.

__

Learning to classify coffee from cappuccino - Part 2

February 24, 2019

This post implements a logistic regression algorithm from scratch with Python and Numpy, using gradient descent and Iteratively reweighted least squares (IRLS) methods.

__

Learning to classify coffee from cappuccino - Part 1

February 17, 2019

This post introduces the logistic regression theory, how to define a probabilistic model of a classification problem and the cross-entropy loss via maximum likelihood.

__

Hello world for Machine learning - Part 5

February 10, 2019

This post explores feature scaling, polynomial features and hypothesis evaluation for linear regression in Scikit-Learn.

__

Hello world for Machine learning - Part 4

February 3, 2019

This post implements a linear regression algorithm in Scikit-Learn and Tensorflow.

__

Hello world for Machine learning - Part 3

January 27, 2019

This post applies the linear regression theory to a multi-input linear case from scratch in Numpy.

__

Hello world for Machine learning - Part 2

January 20, 2019

This post applies the linear regression theory to a single-input case from scratch in Numpy.

__

Hello world for Machine learning - Part 1

January 13, 2019

This post introduces the linear regression as the "hello-world" machine learning problem.

__

Welcome to my personal blog

January 6, 2019

This post introduces you to my blog journey.