Follow along with me and code your own Python Tensorfow Project. In this beginner deep learning classification project with Tensorflow's Sequential model. When we start designing the architecture for a deep learning model it is a good idea to set up a base model so we can evaluate our progress. Sklearn's RandomForestClassifier or GradientBoostingClassifier with the default setting work well as base models.
Random Forest just gives good predictions in a few lines of code and sets the baseline of what we should be trying to beat with our Tensorflow Sequential model.
With Sklearn and Tensorflow we'll predict if employees will stay at the company and help the coporate planning team.
Follow Data Science Teacher Brandyn
dataGroups:
Through many experiments using a simple for loop in Python, we determine the correct architecture for our Sequential model in Tensorflow.
Complete univariate analysis in for loop with Pandas .plot to quickly understand the distributions in your our employement dataset.
Next, plot the univariate distributions conditionally color-coded ny our target. In our deep learning classification workflow, we are predicting if an employee with Leave or Not from their current position.
One hot encoding our categories with pandas .get_dummies to use our categories in our predictive deep learning model.
Train Test Split using Sklearn. Follow by standardization and principal component again with Sklearn.
Use Sklearn's RandomForestClassifier to get our baseline accuracy score.
But a model_factory function that builds our model so we can run through a series of experiments to get our model to work well.
Our deep learning experimental playground to find the best architecture for our Tensorflow model.