Siraj Raval has designed a free curriculum to help anyone learn Computer Vision in the most efficient way possible.

My curriculum starts off with low level vision techniques and progressively increases in difficulty until we get to high level analysis techniques i.e deep learning. Don’t worry if you’ve never coded before, i’ve included links to help you learn Python as well. Now is the time to build computer vision solutions, the world needs these menial tasks automated to help liberate humans from drudgery. The tools needed are python, OpenCV, and Tensorflow, all of which have their place and I’ll explain all the details of how it fits together in this video. Enjoy!

Along the lines of a recent post about companies that still run SQL 2008, here’s a thread from Slashdot (!) about how the quest for “latest and greatest” often over complicates enterprise tech.

“For better or worse, the world runs on Excel, Java 8, and Sharepoint, and I think it’s important for us as technology professionals to remember and be empathetic of that.”

Time series forecasting, Text, and music generation scenarios have made made great strides recently using LSTM Neural Networks. In this article, learn how to train a LSTM Neural Network for text generation in the style of H. P. Lovecraft.

A LSTM (Long Short-term Memory) Neural Network is just another kind of Artificial Neural Network, which falls in the category of Recurrent Neural Networks.

What makes LSTM Neural Networks different from regular Neural Networks is, they have LSTM cells as neurons in some of their layers.

Chatbots act as friendly assistants that make life easier by helping us book flights, appointments, shop, get answers to our questions, etc.  Chatbots are everywhere and coming to more places. Customers get 24/7 support and companies don’t have to spend a fortune on staff.

In fact, 80% of companies want to have some type of chatbot interface implemented by 2020.

But, what are the security best practices around chatbots and what can be done to mitigate any attack surfaces they expose?

Typically, chatbots that are used in industries such as retail, banking, financial services, and travel handle very important data such as credit/debit cards, SSN, bank accounts, and other Sensitive PII (Personally identifiable information). The collection of this type of data is vital for the chatbot to do its job; therefore, chatbots and others digital assistants become an attractive target to be exploited by an attacker to steal users’ information.

Here’s an interesting look at weight agnostic neural networks and what problems they solve. Very interesting read.

Not all neural network architectures are created equal, some perform much better than others for certain tasks. But how important are the weight parameters of a neural network compared to its architecture? In this work, we question to what extent neural network architectures alone, without learning any weight parameters, can encode solutions for a given task.

Just imagine where AI was just 5 years ago. Sure, neural networks have been around for decades, but they were not practical for the average business problem. Think of all the breakthroughs in machine learning, natural language processing, knowledge graphs, and more just in 2019.

Here’s an interesting report, aptly titled State of AI Report 2019 published on June 28. In it, Benaich and Hogarth embark on a 136-slide long journey on all things AI. From technology breakthroughs and their capabilities to supply, demand and concentration of talent working in the field. There are even special sections on the politics of AI and AI in China.

The report lives up to Benaich’s goals as set in his reply. The first 40 pages of the report, which comes in the shape of a slide deck, are focused on progress in AI research — technology breakthroughs and their capabilities. Key areas covered are reinforcement learning, applications in games and future directions, natural language processing breakthroughs, deep learning in medicine, and AutoML.

The Data Science & AI community has truly embraced open source. In fact, there are so many libraries and tools out there, that it can be challenging to keep up. Fortunately, here’s a great round up of 21 open source tools for Machine Learning. Some of them you may have heard of, but I guarantee there are a few surprises in this list, even for the seasoned expert.

  • Presenting 21 open source tools for Machine Learning you might not have come across
  • Each open-source tool here adds a different aspect to a data scientist’s repertoire
  • Our focus is primarily on tools for five machine learning aspects – for non-programmers(Ludwig, Orange, KNIME), model deployment(CoreML, Tensorflow.js), Big Data(Hadoop, Spark), Computer Vision(SimpleCV), NLP(StanfordNLP), Audio, and Reinforcement Learning(OpenAI Gym)