There are different kinds of data. Neural networks architectures are of different kinds too.
Some of the choices of network architecture can be obvious, but machine learning is an experimentation science. A network that was invented to process images can turn out to work well on texts too.
As of today, neural networks are of 4 main architectures: Densely connected neural networks, Convolutional Neural Networks(Convnets), Recurrent Neural Networks (RNNs), and transformers.
What are these networks and what kinds of data they are suited for?
Densely connected networks are made of stacks of layers from the input to the output.
For more than three months, I have been writing consistently every week on my blog and Medium about machine learning ideas. This week, I want to share my thoughts on how someone can get started with machine learning.
We are fortunate to have many and freely available learning resources, but most of them won’t help because they skip the fundamentals or start with moonshots.
How does someone get started?
The first step to learning a hard topic is to get excited first.
Machine learning is a demanding field and it will take time to start understanding concepts and connecting things.
Decision trees are one of the powerful machine learning algorithms that can be used to model complex datasets. They can be used in both regression and classification problems.
In this article, I want to highlight 7 interesting facts that this kind of learning algorithm is very particular about.
Other than most machine learning models which are black boxes, decision trees are white boxes. The predictions made by decision trees are explainable.
Because trees make it possible to visualize the prediction process in the flow chart format, it’s pretty easy to interpret the results. …
Machine Learning models are very particular about the type and range of values that have to go for their input in order to do what they do.
With the exception of decision trees, most ML models will expect you to scale the input features. When the input features are scaled, the model can converge faster than it would without scaling the input features. What is feature scaling?
Feature scaling is nothing other than transforming the numerical features into a small range of values. Its name implies that only features are scaled. Labels or output data don’t need to be scaled.
If you’re like me, most of the time you’re confused by reading the confusion matrix. After all, it was named after such truth.
A confusion matrix is a table used for displaying the number of samples (in classification problems) that are correctly and incorrectly classified in positive and negative classes.
While a confusion matrix can help us to calculate classification metrics (like accuracy, precision, or recall, and f1 score), it is not itself a metric.
Before we see how to read this kind of matrix, let’s see what it is made of.
A confusion matrix is made of 4 main…
The whole goal of using machine learning is to learn the rules that can be used to automate a given task. Without the data and ability to learn the patterns hidden in the data(inexplicitly), machine learning would not be the headlines of the internet nowadays.
In order for a machine to learn such patterns, we need three things:
Let’s talk about each requirement.
Data is the primary input for any machine learning algorithm.
Data can be structured, or unstructured. Structured data are usually in spreadsheet(tabular) format. …
Machine learning systems are complicated. And sometimes, it’s not the fault of the engineers who build them. It’s the nature of machine learning systems.
Here is what I mean…
Let’s say that you did a great job at finding good data, you prepared it reasonably well, and your model made great predictions. Everything is pretty cool at the moment!
But there are times you won’t be able to prevent the worse to happen. A model that is used to make good predictions can start to make misleading predictions. What can go wrong?
There are two reasons why that can happen…
Recently, I had the fortune to host a world-class Machine Learning practitioner, who has not only built a wide range of Machine Learning systems but also helped many people to make a career in Machine Learning.
He is Santiago Valdarrama, someone you might know if you are on the Twitter ML community. Being a Director of Computer Vision at Levatas, he leads a team of software developers and machine learning engineers in the development of Levatas’ flagship product. …
Exploratory Data Analysis or what many people call EDA is a critical step in a machine learning or data science project. This step is more about learning the data and when done properly, you can find some interesting insights that can help you in understanding why certain predictions were made.
In this post, I want to share how I approach the exploratory analysis. I usually start with simple things that you may already know, especially if you have done any end to end project.
This is the initial step in the process of exploratory analysis. It is here that you…
Addressing the common hidden mispractices that can hurt the results of the machine learning systems when everything else was done well!
In almost any step involved in building a machine learning project, there is a chance that something can be done incorrectly. There is going to be a small mispractice that is hard to notice but can completely ruin everything.
Here is what I want to mean…
The failure of a machine learning project can be caused by many factors but the two common pitfalls are data leakage and inconsistent data preprocessing functions. …