The architecture behind Michelangelo uses a modern but complex stack based on technologies such as HDFS, Spark, Samza, Cassandra, MLLib, XGBoost, and TensorFlow.

Artificial Intelligence is spreading its wings almost everywhere. Starting from the businesses to even the agricultural fields, AI is powering the world in many ways than one. There have been various discussions surrounding the fact that AI has the potential to impact the education sector as well. There seems to be various possibilities that Artificial Intelligence is expected to spur innovation in the field of education. We use education as a way to create minds equipped with utilizing and expanding the knowledge pool, whereas Artificial Intelligence provides tools for building up a progressively exact and comprehensive picture of how the human brain performs.

Ever fantasied about having your own personal assistant to answer any questions you can ask, or have conversations with? Well, thanks to Machine Learning and Deep Neural Networks, this is not so far from happening. Think of the amazing capabilities exhibited by Apple's Siri or Amazon's Alexa. Don't get too excited, in this next series of posts we are not going to create an omnipotent Artificial Intelligence, rather we will create a simple chatbot that given some input information and a question about such information, responds to yes/no questions regarding what it has been told. It is nowhere near to Siri's or Alexa's capabilities, but it illustrates very well how even using very simple deep neural network structures, amazing results can be obtained.

To cooperate with humans effectively, virtual agents need to be able to understand and execute language instructions. A typical setup to achieve this is with a scripted teacher which guides a virtual agent using language instructions. However, such setup has clear limitations in scalability and, more importantly, it is not interactive. Here, we introduce an autonomous agent that uses discrete communication to interactively guide other agents to navigate and act on a simulated environment. The developed communication protocol is trainable, emergent and requires no additional supervision. The emergent language speeds up learning of new agents, it generalizes across incrementally more difficult tasks and, contrary to most other emergent languages, it is highly interpretable. We demonstrate how the emitted messages correlate with particular actions and observations, and how new agents become less dependent on this guidance as training progresses. By exploiting the correlations identified in our analysis, we manage to successfully address the agents in their own language.

Deep Reinforcement Learning (DRL) has become a powerful methodology to solve complex decision-making problems. However, DRL has several limitations when used in real-world problems (e.g., robotics applications). For instance, long training times are required and cannot be accelerated in contrast to simulated environments, and reward functions may be hard to specify/model and/or to compute. Moreover, the transfer of policies learned in a simulator to the real-world has limitations (reality gap). On the other hand, machine learning methods that rely on the transfer of human knowledge to an agent have shown to be time efficient for obtaining well performing policies and do not require a reward function. In this context, we analyze the use of human corrective feedback during task execution to learn policies with high-dimensional state spaces, by using the D-COACH framework, and we propose new variants of this framework. D-COACH is a Deep Learning based extension of COACH (COrrective Advice Communicated by Humans), where humans are able to shape policies through corrective advice. The enhanced version of D-COACH, which is proposed in this paper, largely reduces the time and effort of a human for training a policy. Experimental results validate the efficiency of the D-COACH framework in three different problems (simulated and with real robots), and show that its enhanced version reduces the human training effort considerably, and makes it feasible to learn policies within periods of time in which a DRL agent do not reach any improvement.

For over 50 years, researchers have been trying to teach computers to read music notation, referred to as Optical Music Recognition (OMR). However, this field is still difficult to access for new researchers, especially those without a significant musical background: few introductory materials are available, and furthermore the field has struggled with defining itself and building a shared terminology. In this tutorial, we address these shortcomings by (1) providing a robust definition of OMR and its relationship to related fields, (2) analyzing how OMR inverts the music encoding process to recover the musical notation and the musical semantics from documents, (3) proposing a taxonomy of OMR, with most notably a novel taxonomy of applications. Additionally, we discuss how deep learning affects modern OMR research, as opposed to the traditional pipeline. Based on this work, the reader should be able to attain a basic understanding of OMR: its objectives, its inherent structure, its relationship to other fields, the state of the art, and the research opportunities it affords.

Thanks to new technology that enables them to create simple tools, robots may be on the verge of their own version of the Stone Age. Using a novel capability to reason about shape, function, and attachment of unrelated parts, researchers have for the first time successfully trained an intelligent agent to create basic tools by combining objects. The concept may sound familiar. It's called "MacGyvering," based off the name of a 1980s--and recently rebooted--television series. In the series, the title character is known for his unconventional problem-solving ability using differing resources available to him.

Summer is winding down and that means the return to the classroom just around the corner, however, the school supply list isn't the only way to prep for the upcoming academic year. Outfitting your residence with smart home gadgets can make the busy back-to-school-season easier for the whole family. Does your student walk home from school? A digital smart lock can be a lifesaver if your child ever loses his or her house keys. Don't want your child fiddling with the air conditioner after coming home from soccer practice?

This paper presents a physics-based data-driven method to learn predictive reduced-order models (ROMs) from high-fidelity simulations, and illustrates it in the challenging context of a single-injector combustion process. The method combines the perspectives of model reduction and machine learning. Model reduction brings in the physics of the problem, constraining the ROM predictions to lie on a subspace defined by the governing equations. This is achieved by defining the ROM in proper orthogonal decomposition (POD) coordinates, which embed the rich physics information contained in solution snapshots of a high-fidelity computational fluid dynamics (CFD) model. The machine learning perspective brings the flexibility to use transformed physical variables to define the POD basis. This is in contrast to traditional model reduction approaches that are constrained to use the physical variables of the high-fidelity code. Combining the two perspectives, the approach identifies a set of transformed physical variables that expose quadratic structure in the combustion governing equations and learns a quadratic ROM from transformed snapshot data. This learning does not require access to the high-fidelity model implementation. Numerical experiments show that the ROM accurately predicts temperature, pressure, velocity, species concentrations, and the limit-cycle amplitude, with speedups of more than five orders of magnitude over high-fidelity models. Moreover, ROM-predicted pressure traces accurately match the phase of the pressure signal and yield good approximations of the limit-cycle amplitude.

Natural language understanding (NLU) of text is a fundamental challenge in AI, and it has received significant attention throughout the history of NLP research. This primary goal has been studied under different tasks, such as Question Answering (QA) and Textual Entailment (TE). In this thesis, we investigate the NLU problem through the QA task and focus on the aspects that make it a challenge for the current state-of-the-art technology. This thesis is organized into three main parts: In the first part, we explore multiple formalisms to improve existing machine comprehension systems. We propose a formulation for abductive reasoning in natural language and show its effectiveness, especially in domains with limited training data. Additionally, to help reasoning systems cope with irrelevant or redundant information, we create a supervised approach to learn and detect the essential terms in questions. In the second part, we propose two new challenge datasets. In particular, we create two datasets of natural language questions where (i) the first one requires reasoning over multiple sentences; (ii) the second one requires temporal common sense reasoning. We hope that the two proposed datasets will motivate the field to address more complex problems. In the final part, we present the first formal framework for multi-step reasoning algorithms, in the presence of a few important properties of language use, such as incompleteness, ambiguity, etc. We apply this framework to prove fundamental limitations for reasoning algorithms. These theoretical results provide extra intuition into the existing empirical evidence in the field.