All the code and online demo are available at the project page. Describing an image is easy for humans, and we are able to do it from a very young age. In machine learning, this task is a discriminative classification/regression problem, i.e. predicting feature labels from input images. Recent advancements in ML/AI techniques, especially deep learning models, are beginning to excel in these tasks, sometimes reaching or exceeding human performance, as is demonstrated in scenarios like visual object recognition (e.g. from AlexNet to ResNet on ImageNet classification) and object detection/segmentation (e.g. from RCNN to YOLO on COCO dataset), etc. However, the other way around, generating realistic images based on descriptions, is much harder, and takes years of graphic design training.

As Machine Learning (ML) is becoming an important part of every industry, the demand for Machine Learning Engineers (MLE) has grown dramatically. MLEs combine machine learning skills with software engineering knowhow to find high-performing models for a given application and handle the implementation challenges that come up -- from building out training infrastructure to preparing models for deployment. New online resources have sprouted in parallel to train engineers to build ML models and solve the various software challenges encountered. However, one of the most common hurdles with new ML teams is maintaining the same level of forward progress that engineers are accustomed to with traditional software engineering. The most pressing reason for this challenge is that the process of developing new ML models is highly uncertain at the outset.

In my previous post, we discussed the foundation of community detection using modularity optimization. One major constraint however, is that your graph needs to fit in memory. This quickly turns problematic as your number of nodes surpass billions, and the number of edges becomes trillions. Thankfully we can leverage distributed computation systems in order to solve this limitation. To do this we first need to define the state of a node so that it contains all the information needed during computation; this will serve as a basic structure to pass around between the machines of our distributed cluster.

Recently, I gave a talk at the O'Reilly AI conference in Beijing about some of the interesting lessons we've learned in the world of NLP. While there, I was lucky enough to attend a tutorial on Deep Reinforcement Learning (Deep RL) from scratch by Unity Technologies. I thought that the session, led by Arthur Juliani, was extremely informative and wanted to share some big takeaways below. In our conversations with companies, we've seen a rise of interesting Deep RL applications, tools and results. In parallel, the inner workings and applications of Deep RL, such as AlphaGo pictured above, can often seem esoteric and hard to understand.

A successful product has consistent behavior, meets or exceeds user-expectations, and significantly contributes to the top-line growth for the business. It is vital for a Product Manager to set and manage the expectations of users, gather quantifiable feedback regularly, communicate it rigorously to engineers, and make sure the product pragmatically evolves with the business and market transitions. AI products, however, can differ significantly from traditional products. For example, in my prior experience as a Product Manager, success was measured through delivery of a'deterministic' product that always delighted customers -- a hardware product has the same behavior under the standard conditions, the same user actions in a software product results in the same expected response. An AI-driven product, however, may not always have a deterministic behavior and may in fact produce counter-intuitive results -- a personalized recommender system may produce different results to a user action after learning additional preferences.

Video game/E-sports streaming is a huge and ever rising market. In the world championship of League of Legends (LoL) last year, one semifinal attracted 106 million viewers, even more than the 2018 Super Bowl. Another successful example is Twitch, where thousands of players broadcast their gameplay to millions of viewers. Visor, a company that provides personalized game analytics to players, wants a model to estimate the winning rate of a team in real time. There can be many use cases for the model.

For more content like this, follow Insight and Emmanuel on Twitter. When trying to develop a scientific understanding of the world, most fields start with broad strokes before exploring important details. In Physics for example, we start with simple models (Newtonian physics) and progressively dive into more complex ones (Relativity) as we learn which of our initial assumptions were wrong. This allows us to solve problems efficiently, by reasoning at the simplest useful level. The exact same approach of starting with a very simple model can be applied to machine learning engineering, and it usually proves very valuable. In fact, after seeing hundreds of projects go from ideation to finished products at Insight, we found that starting with a simple model as a baseline consistently led to a better end product.

After helping over 1200 Insight Fellows become data scientists, data engineers and AI engineers across the US, we're excited to announce that we're expanding our efforts to enable top product managers to transition to data product management roles. The Insight Data Product Management Fellows Program will start this summer in Silicon Valley and provide experienced PMs with a passion for data an avenue to join top company building AI-enabled products as a Data PM. This tuition-free, full-time, 7-week fellowship is ideally suited for existing product managers, as well as product-focused data scientists, engineers, founders and MBAs with a strong technology background and desire to work as a Data PM. Insight Data PM Fellows will spend the fellowship building products with a team of data scientists, data engineers and AI engineers, while meeting leaders in data and product functions from across Silicon Valley, including teams from LinkedIn, Lyft, Yelp, eBay, Facebook, Salesforce, GitHub, Uber, Google, Cloudera, Box, and Twitter. At the end of the fellowship they will interview with top teams hiring for data product management roles.

Tasks that humans take for granted are often difficult for machines to complete. That's why when you're asked to prove yourself human through those CAPTCHA tests, you're always asked a ridiculously simple question, e.g., whether an image contains a road sign or not, or selecting a subset of images that contain food (see Moravec's Paradox). These tests are effective in determining whether a user is human precisely because image recognition in context is difficult for machines. Training computers to accurately answer these kinds of questions in an automated, efficient way for large amounts of data is complicated. To get around this, companies like Facebook and Amazon spend a lot of money to manually deal with image and video classification problems.

With global climate change, devastating hurricanes are occurring with higher frequency. After a hurricane, roads are often flooded or washed out, making them treacherous for motorists. According to The Weather Channel, almost two of every three U.S. flash flood deaths from 1995–2010, excluding fatalities from Hurricane Katrina, occurred in vehicles. During my Insight A.I. Fellowship, I designed a system that detects flooded roads and created an interactive map app. Using state of the art computer vision deep learning methods, the system automatically annotates flooded, washed out, or otherwise severely damaged roads from satellite imagery.