We are excited to announce the addition of TensorFlow to the Loominus machine learning Provider Plugin architecture! Loominus has taken the complexity out of hand coding TensorFlow models and made it easier for anyone to experiment with neural networks on classification and regression problems. In this post we’ll discuss how to use TensorFlow for Classification and Regression models in Learner.Read More
He demonstrates how to create a data staging and reporting table. (Have a look at how to do feature engineering in Teraport.) Then he trains and compares binary classification models from Scikit-Learn, XGBoost and LightGBM. He does this using Learner’s unified Provider Plugin architecture.
Finally, Hung selects and deploys the best model to an API endpoint using Modops. He goes on to demonstrate how you can use the API in your own applications to get predictions from the model.
One of the most common mistakes data scientists make when training machine learning models is incorrectly splitting data for training and testing. The train/test split involves splitting data during the model training and evaluation process. Usually data is divided into two parts:
- Training data set – The data used to train the model
- Testing data set – The hold out data used to test the performance of the model
Typically we reserve 70% of the data for training and 30% percent for testing. This can vary and should be adjusted depending on the volume of data, the kind of models under consideration and the purpose for modeling.Read More