How to generate the optimal version of your model. — Hyper-Parameter Optimization is a difficult task. However, it can be made easier with tools like Optuna. In this post, I show how to tune the hyper-parameters of a CatBoost model using Optuna. The code here for Optuna can be quickly adapted to whatever model…

A natural selection process to determine an optimal set of features. — A genetic algorithm is a technique for optimization problems based on natural selection. In this post, I show how to use genetic algorithms for feature selection. While there are many well-known feature selections methods in scikit-learn, feature selection goes well beyond what is available there. Feature selection is a crucial…

Quickly set up CRUD operations for MongoDB in python — MongoDB is a no-SQL database popular for its document-oriented database style. If you know what a JSON file or a python dictionary looks like you’ll be familiar with MongoDB documents. Break free from rigid relational database schema with MongoDB. The document style collections (equivalent to tables) enable developers to rapidly…

How to connect to the Google Translate API directly with Python — Here is a brief introduction to using the Google Translate API in python. The Google Translate API is easy to access in python with the help of the package googletrans. This library is easy to use and allows a user to translate python strings from any source language to any…

Small changes which drastically impact model performance — Developing deep learning models requires a great deal of fine-tuning. This article discusses and compares the effects of different activation functions and weight initializers on model performance. This article will cover three activation functions: sigmoid, hyperbolic tangent (tanh), rectified linear unit (ReLU). These activations functions are then tested with the…

How to make good models great through optimization — Deep learning models are incredibly flexible, but a great deal of care is required to make them effective. The choice of learning rate is crucial. This article is the first in a series about fine-tuning deep learning models. This article will discuss the effects of a learning rate on the…

Learn how to understand your machine learning models better. — Machine learning is complicated. There are countless options available and a lot to track. Fortunately, there’s TensorBoard, which makes the process easy. When developing machine learning models, there are many factors: How many epochs for training, the loss metric, or even the model structure. Each of these decisions can propagate…

Create dashboards in Dash to deliver value from your analysis. — Dashboards are a quick and effective way to visualize and monitor your data. With Plotly and Dash, building dashboards in Python has never been more straightforward. This article will discuss why dashboards are helpful and show how to generate a basic dashboard using Dash…