Machine Learning without programming knowledge
Machine Learning is an important technology for handling data in today’s world. It is used to derive models of reality from data. For example, you can use it to segment customer data in an online store or to optimize a performance marketing campaign.
This usually requires the use of a programming language with a large number of program libraries for the selected language. Very often “Python” or “R” are used here today and libraries like “Scikit Learn” and “TensorFlow”.
But there is another way!
Another way the platform “BigML” tries to go is by offering a user interface that allows them to control all steps of a “Machine Learning” project via menus.
“BigML” is a commercial product that can be accessed at https://bigml.com/ and offers different licenses. Besides a free restricted “Personal Account” there is the “Lite” version for 10.000$ per year and the “Enterprise” license for 45.000$ per year. Free licenses are available for students and teachers.
The workflow starts with the import of data from various sources and also provides a quick overview of the data types, correlations, and statistical distributions of characteristics and target variables.
After pre-processing steps, such as the treatment of missing data, have been performed and the data has been split into training and test data, models can be calculated with the different model classes. Here, “BigML” for “Supervised Learning” offers the most common procedures for regression and classification tasks, such as decision trees or “Random Forests”, linear regression or “Logistic Regression” and neural networks.
The most common procedures can be performed with the system
Various methods for “Unsupervised Learning”, such as “cluster analysis”, anomaly detection, and principal component decomposition are also found in BigML’s “toolbox”.
With the models created, predictions can then be made and the quality of the models can be determined with test data. Once a satisfactory model has been assembled, it can be downloaded in the form of program codes for different programming languages and integrated into software solutions, for example.
In order to get through the workflow without having to set a large number of parameters, even with little previous knowledge, “1-Click” wizards are offered in many places in the program interface.
The platform offers a variety of methods that can be used very conveniently and time-saving. However, without a correspondingly good previous knowledge in the area of “Machine Learning”, one will not achieve good results here either. I see a big advantage of the platform in the fact that one is guided through the process in a well-structured way and that many visualizations facilitate the interaction with the software and the understanding of the data.
A short introduction can be found for example at: