I found out at the beginning of my senior year at Berkeley that I would, somehow, complete every single one of my major requirements by the end of my first semester. But I was determined to walk with my friends in the Spring commencement, so the best way I could think of to fill my time was to pick up a minor. This is all to say that I took four data science classes in one semester, spent around 60 hours every week trying to figure out the complicated math and code behind linear and logistic regression, how to accurately model a time series without too much bias or variance, and the magic of how to model a data set accurately so that it would predict the future!
I discovered EazyML recently. The feelings I have for this automated machine learning platform are bilateral: I’m so thankful and impressed with the ease of use and intuitiveness of the step-by-step program, but also annoyed that I spent several hundred hours of my time trying to figure out how to do what EazyML has done for me in ten minutes.
The dataset I chose to put into EazyML was month by month data for COVID-19 cases for every state from January 1 until June 26. I chose this dataset because COVID-19 is the foremost topic on my mind, and I assume, most people’s minds. I imagined that the predictive feature EazyML’s automated AI Platform offers would be a great way to forecast how US cases and deaths would change over the next few months – if accurate, this could potentially help with the uncertainty in every single sector, from Wall Street to academia to everyday life. The Novel Coronavirus 2019 Dataset, can be found here.
Detailed mode is extremely beginner-friendly (only a few Google searches needed), but also provides all the logical information a data scientist needs to understand what the program is doing and how the pipes are functioning in the background. For example, if you are unfamiliar with data science concepts such as feature engineering for machine learning or regression, it would only take you half an hour to brush up on the basics using open-source information to understand the steps EazyML follows (Why EazyML?).
As a beginner, I simply upload any dataset – for me, COVID-19 cases and deaths by state over time – then answer a few simple questions about the dataset, wait a minute or so for EazyML to build models it would have taken me 5 hours to build, and then relax and enjoy the final model that EazyML outputs with confidence that the model it chose has the lowest error. In my case, this AI platform suggested Neural Network Regression with an r-squared value of only 0.25. Then, after just two or three clicks, I obtain the results of the model’s predictions and if curious, can ask the platform to explain to me the results of the prediction. One of the features I appreciate the most is the visualizations along the way for you to understand your dataset further. This makes it so that you not only complete your task but also learn about data science along the way. We see that the number of COVID cases are trending down, albeit slowly. I believe the data failed to account for a second wave that we are seeing now. If variables were held constant, such as the stay-at-home order and mask regulations, the predictions would have likely been a much more accurate depiction of what we’re experiencing today. In this sense, machine learning could serve as a good lesson to our politicians – using automated machine learning platforms like EazyML to drive policy may be more effective than political advice and conjecture.
EazyML makes it easy for people with little to no data science experience to get the most out of their data (Growth of Automated Machine Learning). As long as you understand your dataset, EazyML makes it incredibly simple for every person to make the most out of their data, from analysts in enterprise to students of any age looking to advance beyond their coursework or pick up a new skill. Say you’re running a business and you want to perform machine learning on your datasets. Your options are to hire a team of data scientists, or go to EazyML and save yourself hundreds of hours and thousands upon thousands of dollars. Say you’re a current college student not studying data or computer science, but want to know the future of computing. EazyML can save you countless hours and tuition fees by analyzing your data set for you and teaching you about data science along the way. Say you’re stuck in a job you aren’t necessarily passionate about. EazyML is the simplest way to explore your interest in a future career in machine learning for the lowest cost and least time possible.
Some of the downfalls of EazyML lie in the details. Some of the datasets that EazyML provides itself cannot be used in EazyML because the training data is fewer than 100 lines. The UI could be more modern – the buttons remind me of software from the early 2000s. I hope EazyML creators read this article of mine and make necessary enhancements. Let’s have a look.
What differentiates EazyML from the other automated AI platforms on the market is most markedly its ease of use (Democratizing And Improving AI). With other platforms, you must either be part of an enterprise or organization to access all features and you must have some fundamental knowledge of the systems and functions of computer programming and data science in order to maximize the potential of these platforms. This is not the case for EazyML. Simply, EazyML has the lowest barrier to entry for any automated AI platform currently on the market. Microsoft Azure, though incredibly competent, is extremely complicated with too many features to easily comprehend for an individual. Amazon SageMaker requires background competency with AWS and requires an expensive fee for their entire suite of applications. If you only need to perform machine learning tasks without paying for another slew of functions that you have no need for, EazyML is the only and best option.
As machine learning becomes more and more integrated into our everyday lives, the majority of data-unsavvy people will become more and more confused as to how the apps, websites, and news we consume every day are working behind the scenes. EazyML alleviates this confusion for the general public and empowers the everyday man to harness the power of machine learning for use in their everyday life – easily.