Source: Data Science Central | 2019
Author: Divya Singh
Trends to Watch Out For and Prepare Yourself
Ok. So there’s been a lot of coverage by various websites, data science gurus, and AI experts about what 2019 holds in store for us. Everywhere you look, we have new fads and concepts for the new year. This article is going to be rather different. We are going to highlight the dark horses – the trends that no one has thought about but will completely disrupt the working IT environment (for both good and bad – depends upon which side of the disruption you are on), in a significant manner. So, in order to give you a taste of what’s coming up, let’s go through the top four (plus 1 (bonus) = five) top trends of 2019 for data science:
- Interoperability (ONNX)
- Cyber Data Science Crime
- Cloud AI-as-a-Service
- (Bonus) Quantum Computation & Data Science
1. AutoML (& AutoKeras)
This single innovation is going to change the way machine learning works in the real world. Earlier, deep learning and even advanced machine learning was an aristocratic property of PhD holders and other research scientists. AutoML has changed that entire domain – especially now that AutoKeras is out.
AutoML automates machine learning. It chooses the best architecture by analyzing the data – through a technology called Neural Architecture Search (NAS), tries out various models and gives you the best possible hyperparameters for your scenario automatically! Now, this was priced at the ridiculous price of 76$ USD per hour, but we now have a free open source competitor, AutoKeras.
AutoKeras is an open source free alternative to AutoML developed at University of Texas A & M DATA lab and the open source community. This project should make a lot of deep learning accessible to everyone on the planet who can code even a little. To give you an example, this is the code used to train an arbitrary image classifier with deep learning:
import autokeras as ak clf = ak.ImageClassifier() clf.fit(x_train, y_train) results = clf.predict(x_test )
Folks, it really doesn’t get simpler than this!
Note:Of course, the entire training and testing process will take more than a day to complete at the very least, but less if you have some high-throughput GPUs or Google’s TPUs (Tensor Processing Units – custom hardware for data science computation) or plenty of money to spend on the cloud infrastructure computation resources of AutoML.