Artificial intelligence (AI) has certainly become a popular buzzword. Today, and every day for the foreseeable future, a multitude of industries and companies are contemplating how to incorporate AI into their operations. To implement such solutions into your company, though, answering a few questions about feasibility is essential. To understand if your company is ready to utilize artificial intelligence, you must first ask why you will be employing this technology. What problem are you trying to solve?
After some frustration, I ended up with a patchy solution that does the work for me. It's not the nicest thing, but works regardless of how you reference your Keras model. Basically, if an object has __getstate__ and __setstate__ methods, pickle will use them to serialize the object. The problem is that Keras Model doesn't implement these. I have a general python module that I always import on all my notebooks and contains some stuff I always need so I just added it there.
Machine Learning is an application of Artificial Intelligence and is revolutionizing the way companies execute business. At its heart, it's an algorithm or model that learns patterns in big data and then predicts synonymous patterns in new data. In layman's terms, it's about machines making sense out of data in much the same way that we humans do. Machine learning helps improve our lives in a huge number of ways. Some of them so ingrained into our lives, life would be unimaginable without them!
In the course of my work as a consultant for Dell Technologies Customer Solution Centers, I work with many business and IT leaders who want to capitalize on the opportunities brought by big data and advances in artificial intelligence. In this blog post, I will walk through some top-of-mind considerations for organizations that are embarking on the journey to AI. What are the top considerations for the move from predictive analytics to AI? The answer is really in the question here. The move to AI should be just that -- a move, rather than an abrupt implementation of AI alone.
Detailed routing is one of the most critical steps in analog circuit design. Complete routing has become increasingly more challenging in advanced node analog circuits, making advances in efficient automatic routers ever more necessary. In this work, we propose a machine learning driven method for solving the track-assignment detailed routing problem for advanced node analog circuits. Our approach adopts an attention-based reinforcement learning (RL) policy model. Our main insight and advancement over this RL model is the use of supervision as a way to leverage solutions generated by a conventional genetic algorithm (GA). For this, our approach minimizes the Kullback-Leibler divergence loss between the output from the RL policy model and a solution distribution obtained from the genetic solver. The key advantage of this approach is that the router can learn a policy in an offline setting with supervision, while improving the run-time performance nearly 100x over the genetic solver. Moreover, the quality of the solutions our approach produces matches well with those generated by GA. We show that especially for complex problems, our supervised RL method provides good quality solution similar to conventional attention-based RL without comprising run time performance. The ability to learn from example designs and train the router to get similar solutions with orders of magnitude run-time improvement can impact the design flow dramatically, potentially enabling increased design exploration and routability-driven placement.