In this tutorial, we’ll show you what it takes to be practical with Machine Learning. First, we’ll show you how to build a working environment for machine learning. Then you will immediately start working with your first machine learning model.
First, you will create a working environment with the following components:
Of course, you can reuse this working environment in the other models of our portfolio.
Then you will work in this environment with a basic Image Classification model. Let’s start!
Setting Up the Working Environment
First, you have to put together your working environment. This takes three steps:
- Install Python, if necessary (it is pre-installed on most computers).
- Install the necessary Python packages.
- Install Jupyter notebook.
After these 3 steps you will have a fully functional working environment for working with machine learning models.
Installation of Python
Python is preinstalled on most computers. You do not have to do anything here. In case it is not installed on your machine, you can download it from Python.org. We recommend to use at least Python 3.5.
Installation of the Python Packages
Once Python is installed, you can install the necessary Python packages. Open a console and execute the following commands:
pip3 install --upgrade pip pip3 install tensorflow pip3 install keras pip3 install numpy pip3 install pandas pip3 install mathplotlib pip3 install pickle
If you are working with Python 2, use pip instead of pip3.
Installation of Jupyter Notebook
To install jupyter notebook, run the following command in the console:
pip3 install jupyter
More information can be found on the Jupyter website.
You now have a complete working environment and can start working.
Loading the Image Classification Model
To load the first example and display it in Jupyter Notebook, follow these steps:
- Create a working directory.
- Download the Image Classification Model and unpack it in the created working directory.
- Start the Jupyter Notebook.
Creating a Working Directory
Create a working directory for the example models and change to this directory:
mkdir ipublia cd ipublia
Downloading the Image Classification Model
Download the Image Classification Model and unpack the downloaded zip file in your working directory (e.g. ~/ipublia).
Starting the Jupyter Notebook
Start the Jupyter Notebook in the console:
In your web browser, a page opens with the list of the installed models. Select the model Image Classification With InceptionV3. There you will find more information on how to work with the model.
In our portfolio you will find more pre-trained models.