Archive: 2014-10

The Watson Explorer The Watson Developer Cloud brings Watson to developers and the cognitive cloud to Internet applications. Watson offers a variety of services for building cognitive apps, including language identification and translation, interpreting meaning based on context, and communicating with people in their own styles. Here are some reviews and links to APIs and sample code. IBM's Watson Supercomputer Gives Developers Access to Cognitive Cloud: George Lawton at TechTarget provides an early review of the Watson Explorer’s unified view of enterprise information. The cloud allows the technology to be accessible for a greater variety of applications and improves the scale and time to market of those applications. IBM Debuts First Watson Machine-Learning APIs: Serdar Yegulalp at Java World previews the eight services that developers can access for building cognitive apps based on Watson’s machine intelligence service. He focuses on visualization rendering as the service least limited by data training.

Deep Neural Networks  “At some fundamental level, no one understands machine learning.” —Christopher Olah “Neural networks are one of the most beautiful programming paradigms ever invented.” —Michael Nielsen This week, we round up a few examples on deep neural networks (DNNs), a subfield of machine learning that deals with developing training algorithms and uses raw video and speech data as input. Replicating Deep Mind: Kristjan Korjus is working on a project to reproduce the results of Playing Atari with Deep Reinforcement Learning, by Volodymyr Mnih and colleagues of DeepMind Technologies. Mnih et al. presented a deep learning model that used reinforcement learning to learn control policies from sensory input and outperformed human experts on three of seven Atari games. Deep Learning, NLP, and Representations: Christopher Olah at Colah's Blog looks at deep learning from a perspective on natural-language processing and discusses how different DNNs designed for different language-processing tasks have learned the same things.

We at the SEI are excited about the Team Software Process (TSP) Symposium, which we are holding in Pittsburgh, Pa. November 3-6, 2014. The theme of the symposium is "Going Beyond Methodology to Maximize Performance." By this, we mean that the technical program goes beyond the core methodology of TSP to encompass a broader range of complementary practices that contribute to peak performance on system and software projects. As part of our strategy to expand the scope of the symposium and bring in more architectural thinking to those who have adopted TSP and are using it, we've added several architecture-related sessions to the technical program. We at the SEI have seen how successful combining TSP and architecture-centric engineering approaches can be in the project we undertook with Bursatec, the technology subsidiary of the Mexican stock exchange.