Bidirectional LSTMs are an extension of traditional LSTMs that can improve model performance on sequence classification problems. In problems where all timesteps of the input sequence are available, Bidirectional LSTMs train two instead of one LSTMs on the input sequence. The first on the input sequence as-is and the second on a reversed copy of the input sequence. This can provide additional context to the network and result in faster and even fuller learning on the problem. In this tutorial, you will discover how to develop Bidirectional LSTMs for sequence classification in Python with the Keras deep learning library.

Artificial intelligence, augmented reality and virtual reality are here to stay, but what impact will they have on jobs and training? A new study by Pew Research Center and Elon University's Imagining the Internet Center asked more than 1,400 technologists, futurists and scholars whether well-prepared workers be able to keep up in the race with artificial intelligence tools, and what impact this development will have on market capitalism. According to Elon University, most of the experts said they hope to see education and jobs-training ecosystems shift in the next decade to exploit liberal arts-based critical-thinking-driven curriculums; online courses and training amped up by artificial intelligence, augmented reality and virtual reality; and scaled-up apprenticeships and job mentoring. However, some expressed fears that education will not meet new challenges or -- even if it does -- businesses will implement algorithm-driven solutions to replace people in many millions of jobs, leading to a widening of economic divides and capitalism undermining itself. "The vast majority of these experts wrestled with a foundational question: What is special about human beings that cannot be overtaken by robots and artificial intelligence?" said Lee Rainie, director of internet, science and technology research at Pew Research Center and co-author of the report.

As such, it was only natural for them to decide to enroll in the Japanese language and culture course that Makati Science High School has been offering ninth- and 10th-graders over the past few years. "My reason for joining the nihongo (Japanese-language) class was initially for anime. But halfway through, I realized I can use it to be able to experience going to Japan," 16-year-old Franza, who took the elective course for two years starting in 2015, said in a recent interview. Chee, who is a year younger than Franza, said, "One of my life goals (includes) watching anime without subtitles. The Filipino Department of Education started offering the Japanese language and culture program to high school students in 2009, together with Spanish and French, to prepare young Filipinos for both local and international opportunities that would require communicative competence in a second foreign language, after English. Mandarin Chinese and German have subsequently been added. The foreign language programs are offered in selected schools across the country, with Japanese taught in 38 schools, mostly in Manila. So far, more than 3,000 students, nearly all avid viewers of Japanese anime, have enrolled in the Japanese program while in grades nine and 10. "We have to recognize the fact that the globe is getting smaller and smaller," Education Secretary Leonor Briones said of the program's relevance. Many students want to go on to further their studies, both at the undergraduate and graduate levels, and Japan is one of the preferred overseas destinations, given the availability of many scholarships to study there, the secretary said. "So it helps if they take lessons.

Mumbai-based Wellthy Therapeutics, a digital therapeutics company that uses artificial intelligence (AI) and patient centric design to improve patient outcomes for all healthcare stakeholders in diabetes care, announced results from a pilot designed to evaluate the feasibility and scalability of an artificial intelligence led lifestyle intervention to improve self-management of people living with type 2 diabetes as a supporting tool to existing care in India. The results were featured via an oral presentation at the American Diabetes Association (ADA) 77th Scientific Sessions, taking place at the San Diego Convention Centre in San Diego, California. The pilot results suggest that continuity of care between physician appointments for people with type 2 diabetes can be achieved with positive outcomes in a clinically significant, scalable and affordable way through this program. Participants that completed the pilot on average dropped their average blood sugar levels (HbA1c) by 0.59%. Amongst the participants that completed and dropped their HbA1c, the average drop observed was even higher at 1.04%.

The plan to hold America's first sextech hackathon wasn't hatched by Silicon Valley brogrammers between swigs of beer and writing code. In an origin story far more fitting, the vision crystallized after a sexual education workshop held at a polyamorous, sex positive living communityin Brooklyn. "You were the ideas man and pulled it all together," Bryony Cole, The Future Of Sex podcast host, said to organizer Andriy Yaroshenko the morning of the innagural event, which was held at the ThoughtWorks office in Manhattan. The workshop where Cole and Yaroshenko met was at the Hacienda Villa housing community, which hosts similar events throughout the year, though they previously connected on Twitter. The pair formed a friendship over a mutual desire to change the way we talk and think about sex.

Designing effective exploration-exploitation algorithms in Markov decision processes (MDPs) with large state-action spaces is the main challenge in reinforcement learning (RL). In fact, the learning performance degrades with the number of states and actions in the MDP. However, MDPs often exhibit a low-dimensional latent structure in practice, where a small hidden state is observable through a possibly large number of observations. In this paper, we study the setting of rich-observation Markov decision processes (\richmdp), where hidden states are mapped to observations through an injective mapping, so that an observation can be generated by only one hidden state. While this mapping is unknown a priori, we introduce a spectral decomposition method that consistently estimates how observations are clustered in the hidden states. The estimated clustering is then integrated into an optimistic algorithm for RL (UCRL), which operates on the smaller clustered space. The resulting algorithm proceeds through phases and we show that its per-step regret (i.e., the difference in cumulative reward between the algorithm and the optimal policy) decreases as more observations are clustered together and finally, matches the (ideal) performance of an RL algorithm running directly on the hidden MDP.

Topic models such as Latent Dirichlet Allocation (LDA) have been widely used in information retrieval for tasks ranging from smoothing and feedback methods to tools for exploratory search and discovery. However, classical methods for inferring topic models do not scale up to the massive size of today's publicly available Web-scale data sets. The state-of-the-art approaches rely on custom strategies, implementations and hardware to facilitate their asynchronous, communication-intensive workloads. We present APS-LDA, which integrates state-of-the-art topic modeling with cluster computing frameworks such as Spark using a novel asynchronous parameter server. Advantages of this integration include convenient usage of existing data processing pipelines and eliminating the need for disk writes as data can be kept in memory from start to finish. Our goal is not to outperform highly customized implementations, but to propose a general high-performance topic modeling framework that can easily be used in today's data processing pipelines. We compare APS-LDA to the existing Spark LDA implementations and show that our system can, on a 480-core cluster, process up to 135 times more data and 10 times more topics without sacrificing model quality.

This paper introduces Dex, a reinforcement learning environment toolkit specialized for training and evaluation of continual learning methods as well as general reinforcement learning problems. We also present the novel continual learning method of incremental learning, where a challenging environment is solved using optimal weight initialization learned from first solving a similar easier environment. We show that incremental learning can produce vastly superior results than standard methods by providing a strong baseline method across ten Dex environments. We finally develop a saliency method for qualitative analysis of reinforcement learning, which shows the impact incremental learning has on network attention.

Several important applications, such as streaming PCA and semidefinite programming, involve a large-scale positive-semidefinite (psd) matrix that is presented as a sequence of linear updates. Because of storage limitations, it may only be possible to retain a sketch of the psd matrix. This paper develops a new algorithm for fixed-rank psd approximation from a sketch. The approach combines the Nystrom approximation with a novel mechanism for rank truncation. Theoretical analysis establishes that the proposed method can achieve any prescribed relative error in the Schatten 1-norm and that it exploits the spectral decay of the input matrix. Computer experiments show that the proposed method dominates alternative techniques for fixed-rank psd matrix approximation across a wide range of examples.

Traditional recommendation systems rely on past usage data in order to generate new recommendations. Those approaches fail to generate sensible recommendations for new users and items into the system due to missing information about their past interactions. In this paper, we propose a solution for successfully addressing item-cold start problem which uses model-based approach and recent advances in deep learning. In particular, we use latent factor model for recommendation, and predict the latent factors from item's descriptions using convolutional neural network when they cannot be obtained from usage data. Latent factors obtained by applying matrix factorization to the available usage data are used as ground truth to train the convolutional neural network. To create latent factor representations for the new items, the convolutional neural network uses their textual description. The results from the experiments reveal that the proposed approach significantly outperforms several baseline estimators.