ENGED PROGRAM CONTRIBUTOR

Collins Ayuya

Collins Ayuya is pursuing his Masters in Computer Science, carrying out academic research in Natural Language Processing. He is a startup founder and is passionate about startups, innovation, new technology, and developing new products. Collins enjoys doing pencil and graphite art and is also a sportsman and gamer.

Author Insights

30

0

0

EngEd Community Badges

Peer
Reviewer

Contributed articles

Entropy and Information Gain to Build Decision Trees in Machine Learning

Jul 3, 2021

A decision tree is a supervised learning algorithm used for both classification and regression problems. Simply put, it takes the form of a tree with branches representing the potential answers to a given...

0

0

0
A Primer on Bayesian Optimization to Optimize Hyperparameters

May 11, 2021

In the two previous optimization articles, we carried out basic experiments (these are listed below). These experiments involved building many models with varied hyperparameters. The experiments of the two...

0

0

0
Using Random Search to Optimize Hyperparameters

Mar 30, 2021

There are many ways to optimize the hyperparameters of a model. We previously looked at grid search and a basic example of its implementation.

0

0

0
Using Grid Search to Optimize Hyperparameters

Mar 18, 2021

Unlike parameters, hyperparameters are set before training a machine learning model. These hyperparameters need to be optimized to adapt a model to a dataset. However, the optimal hyperparameter settings on one...

0

0

0
How to Detect and Correct Multicollinearity in Regression Models

Mar 1, 2021

The interpretability of machine learning models helps us understand the predictions made by a model. Interpretable models are desirable. The interpretability of a regression model may be affected, when...

0

0

0
Parametric versus Non-Parametric Models

Feb 22, 2021

In machine learning, a learner/model learns from training data to map a target function. However, the configuration of the function is undetermined. As a result, we test out many machine learning algorithms to...

0

0

0
Problem Framing in Machine Learning

Feb 18, 2021

Choosing a machine learning method to implement data is not the easiest of processes. It is essential to first understand the precise business problem and its objectives. For instance, understanding what needs...

0

0

0
Regularization to Prevent Overfitting

Feb 8, 2021

Overfitting impacts the accuracy of Machine Learning models. The model attempts to capture the data points that do not represent the accurate properties of data. These data points may be considered as noise. To...

0

0

0
Ensemble Learning on Bias and Variance

Jan 20, 2021

In my previous article, we covered ensemble learning and how it improves machine learning. We mentioned that it reduces the errors of bias and variance. These errors can affect the performance of a machine...

0

0

0
Hypothesis Testing in Data Science

Jan 17, 2021

When using machine learning, we need to be able to trust our models and the predictions they make. We may use sample data to train our models. This sample data may make certain assumptions about a population.

0

0

0
An Overview of Meta-Learning

Jan 13, 2021

The application of machine learning to real-world problems is a very involving task for data scientists. For instance, they need to assess algorithms and tune many parameters. This may be based on their...

0

0

0
Minimizing Data Leakage in Machine Learning

Jan 13, 2021

We always want to produce the best performing machine learning models. If a model does well on our test set, we consider it to be a suitable model. Yet, a model may perform well during testing but poorly after...

0

0

0
Reproducibility to Improve Machine Learning

Jan 12, 2021

In machine learning, reproducibility involves getting the same results, using the same algorithm, data, and parameters on every run. It helps to remove variations when rerunning experiments, which helps in...

0

0

0
Addressing Underspecification in Machine Learning

Dec 21, 2020

It is common to have machine learning models that perform well after training but offer a poor performance after being deployed. A possible cause of this is known as underspecification. Models are supposed to...

0

0

0
Introduction to Active Learning

Dec 17, 2020

By their very nature machine learning models are data-hungry. They require a lot of data to make decisions. However, always having high quality labeled data available to train models is often a costly (and...

0

0

0
Ensemble Learning Techniques to Improve Machine Learning

Dec 13, 2020

We build machine learning models to provide the best possible predictions to a given problem. However, a single model may not make the best predictions and may be subject to errors such as variance and bias.

0

0

0
Handling Imbalanced Datasets in Machine Learning

Dec 3, 2020

Imbalanced datasets mean that the number of observations differs for the classes in a classification dataset. This imbalance can lead to inaccurate results.

0

0

0
Correcting Dataset Shift in Machine Learning

Nov 30, 2020

Dataset shift occurs when there is a change in the distribution of data. In the real world, socioeconomic factors may influence data distribution, along with varying consumer habits, and many other factors.

0

0

0
Evaluating Machine Learning Model Performance

Nov 26, 2020

We expect machine learning models to provide accurate and trustworthy predictions. To confidently trust their predictions, it is important to assess how machine learning models generalize on test data. Let us...

0

0

0
Securing AI Systems from Adversaries

Nov 19, 2020

Artificial intelligence represents just how powerful and impactful technology has become. It is present in all aspects of our daily lives, from basic tasks to very critical implementations. However, AI systems...

0

0

0
Machine Teaching to Improve Artificial Intelligence

Nov 17, 2020

We provide machine learning models with volumes of data expecting that they will identify patterns or learn relationships within the data. This can be an expensive and time-consuming process.

0

0

0
Transfer Learning for Deep Learning

Nov 9, 2020

Deep learning models need a lot of data to work properly. They may be trained on millions of data points before making predictions. This makes the training process very expensive. Furthermore, such models often...

0

0

0
Sustainable AI with Tiny Machine Learning

Nov 5, 2020

Tiny machine learning is at an intersection of embedded devices and machine learning. It is being touted as the next AI revolution. However, the carbon footprint of AI has been increasing with the evolution of...

0

0

0
Basics of Transfer Learning

Oct 28, 2020

When we as human beings learn new things, we do not always learn from scratch. We often transfer past knowledge and apply them to new tasks. For example, knowing how to ride a bike makes it easier to learn how...

0

0

0
Explainable AI - An Introduction

Oct 22, 2020

AI-powered systems have a lot of influence on our daily lives. A number of these systems are so sophisticated that little to no human intervention is required in their design and deployment. These systems make...

0

0

0
Feature Engineering in Machine Learning

Oct 19, 2020

The features used in a machine learning model are often the difference between model success, mediocrity, and failure. Therefore, it is not enough to simply build models, but also making sure they offer the...

0

0

0
Introduction to GPT3

Oct 10, 2020

Generative pre-trained transformers are language models that produce human-like text. These models make it easier to develop machine learning applications while also providing a way in for people with little...

0

0

0
Handling Bias in Machine Learning

Oct 8, 2020

We are increasingly dependent on machine learning in our daily lives. An example is in the every-day products we use. However, these algorithms are created by human beings. As a result, algorithms may contain...

0

0

0
Industry 4.0 and Cybersecurity

Sep 22, 2020

Industrial revolutions describe transitional periods that welcome new manufacturing procedures. The first three industrial revolutions can be described as mechanization to improve manufacturing process,...

0

0

0
Introduction to Automated Fake News Detection

Sep 18, 2020

Misinformation presents a huge challenge in online society. As a result, there have been many attempts to identify and classify misinformation. Specifically, in social networking sites, blogs, as well as online...

0

0

0

Want to learn more about the EngEd Program?

Discover Section's community-generated pool of resources from the next generation of engineers.