Auto-Model: Utilizing Research Papers and HPO Techniques to Deal with the CASH problem Chunnan Wang, Hongzhi Wang, Tianyu Mu, Jianzhong Li, Hong Gao Department of Computer Science Harbin Institute of T echnology Harbin, China {WangChunnan, wangzh, mutianyu, lijzh, honggao }@hit.edu.cn Abstract --In many fields, a mass of algorithms with completely different hyperparameters have been developed to address the same type of problems. Choosing the algorithm and hyperpa-rameter setting correctly can promote the overall performance greatly, but users often fail to do so due to the absence of knowledge. How to help users to effectively and quickly select the suitable algorithm and hyperparameter settings for the given task instance is an important research topic nowadays, which is known as the CASH problem. In this paper, we design the Auto-Model approach, which makes full use of known information in the related research paper and introduces hyperparameter optimization techniques, to solve the CASH problem effectively. Auto-Model tremendously reduces the cost of algorithm implementations and hyperparameter configuration space, and thus capable of dealing with the CASH problem efficiently and easily. T o demonstrate the benefit of Auto-Model, we compare it with classical Auto-Weka approach. The experimental results show that our proposed approach can provide superior results and achieves better performance in a short time. Index T erms--Algorithm selection, Hyperparameter optimization, Combined algorithm selection and hyperparameter optimization problem, Auto-Weka, Classification algorithms I. I NTRODUCTION In many fields, such as machine learning, data mining, artificial intelligence and constraint satisfaction, a variety of algorithms and heuristics have been developed to address the same type of problem [1], [2]. Each of these algorithms has its own advantages and disadvantages, and often they are complementary in the sense that one algorithm works well when others fail and vice versa [2]. If we are capable of selecting the algorithm and hyperparameter setting best suited to the task instance, any particular task instance will be well solved, and our ability of dealing with the problem will be improved considerably [3]. However, it is not trivial to achieve this goal. There are a mass of powerful and different algorithms to deal with a certain problem, and these algorithms have completely different hyperparameters, which have great effect on their performance. Even domain experts cannot easily and correctly select the appropriate algorithm with corresponding optimal hyperparameters from such a huge and complex choice space.

WIRE)--CloudMinds Technology Inc. – a global pioneer in cloud artificial intelligence architecture that makes robots and businesses smarter for the benefit of all humanity – will have its revolutionary XR-1 robot interact with guests at the Sprint exhibit (South Hall #1702) at Mobile World Congress Los Angeles, Oct. 22 to 24. XR-1 is one of the first-ever humanoid robots powered by cloud artificial intelligence, commercial Sprint True Mobile 5G and proprietary vision-controlled grasping technology for service robots that also leverages human operator input for constant learning. "Overall, intelligent cloud robots paint the most vibrant picture of how 5G's ultra-low latency, exponentially faster speeds and wider reach can dramatically improve response time and enable a new world of applications," said Bill Huang, founder and CEO of CloudMinds. "With vision-controlled grasping and the ability to perform intricate tasks, the XR-1 simply raises the bar and lays the foundation for an even wider range of intelligent compliant cloud service robots from CloudMinds – from wheeled to two-legged form factors. We are proud to be ushering in a new era of helpful robots for homes and businesses, with an emphasis on the importance of human input."

The Artificial Intelligence Project to Prevent Gender Violence is an initiative from the Deizy Beltran Program to Protect & Empower Women, Women's Rights Clinc at Ankawa International (Operated by The Ankawa Global Group). The Ankawa Global Group is a private company from Peru developing high-tech tools for policy planning and business intelligence, among other services for public and private actors. Through its Action Tank (Ankawa International), it implements social transformation programs good governance, sustainable development, and innovation worldwide. Our social transformation programs have received international praise and recognition. This artificial intelligence project was presented at the Paris Peace Forum 2019.

Techfest is the annual science and technology festival of Indian Institute of Technology Bombay.[1] It also refers to the independent body of students who organize this event along with many other social initiatives and outreach programs around the year. Techfest is known for hosting a variety of events that include competitions, exhibitions, lectures as well as workshops. Started in 1998 with the aim of providing a platform for the Indian student community to develop and showcase their technical prowess, it has now grown into Asia's Largest Science and Technology Festival[2] with a footfall of 1.75 lakhs in its latest edition.[3][4][5] The activities culminate in a grand three-day event in the campus of IIT Bombay which attracts people from all over the World, including students, academia, corporates and the general public.[6] The very first edition of Techfest was in 1998. The underlying spirit of Techfest was "to promote technology and scientific thinking and innovation" a motto that has been followed by every Techfest since. Techfest '98 also set the broad outlines of Techfest in the form of competitions, lectures, workshops, and exhibitions which went on to become a standard feature at every Techfest. Entrepreneurship also made an appearance in the 1999 and 2000 editions. Technoholix--Techfest in the Dark, showcasing technological entertainment at the end of each day as well as the hub of on the spot activities, made their debut during these years. Techfest 2001-2002 saw the incorporation of IIT Bombay's department oriented events like Yantriki, Chemsplash and Last Straw. Students from G H Raisoni College of Engineering got the Engineering Excellence Award for best design.

Ninety per cent of architects will lose their jobs as artificial intelligence takes over the design process, according to designer Sebastian Errazuriz. The New York-based designer made the controversial claim in a series of movies posted on his Instagram account. "I think it's important that architects are warned as soon as possible that 90 per cent of their jobs are at risk," he said. "It's almost impossible for you to compete" with algorithms he said, adding: "The thing is you're not that special." Born in Chile and raised in London, the artist and designer has courted controversy before, most recently by proposing to turn the fire-damaged Notre Dame cathedral in Paris into a rocket launchpad.

This paper shows the susceptibility of spectrogram-based audio classifiers to adversarial attacks and the transferability of such attacks to audio waveforms. Some commonly adversarial attacks to images have been applied to Mel-frequency and short-time Fourier transform spectrograms and such perturbed spectrograms are able to fool a 2D convolutional neural network (CNN) for music genre classification with a high fooling rate and high confidence. Such attacks produce perturbed spectrograms that are visually imperceptible by humans. Experimental results on a dataset of western music have shown that the 2D CNN achieves up to 81.87% of mean accuracy on legitimate examples and such a performance drops to 12.09% on adversarial examples. Furthermore, the audio signals reconstructed from the adversarial spectrograms produce audio waveforms that perceptually resemble the legitimate audio.

Machine Learning systems are vulnerable to adversarial attacks and will highly likely produce incorrect outputs under these attacks. There are white-box and black-box attacks regarding to adversary's access level to the victim learning algorithm. To defend the learning systems from these attacks, existing methods in the speech domain focus on modifying input signals and testing the behaviours of speech recognizers. We, however, formulate the defense as a classification problem and present a strategy for systematically generating adversarial example datasets: one for white-box attacks and one for black-box attacks, containing both adversarial and normal examples. The white-box attack is a gradient-based method on Baidu DeepSpeech with the Mozilla Common Voice database while the black-box attack is a gradient-free method on a deep model-based keyword spotting system with the Google Speech Command dataset. The generated datasets are used to train a proposed Convolutional Neural Network (CNN), together with cepstral features, to detect adversarial examples. Experimental results show that, it is possible to accurately distinct between adversarial and normal examples for known attacks, in both single-condition and multi-condition training settings, while the performance degrades dramatically for unknown attacks. The adversarial datasets and the source code are made publicly available.

Generative adversarial networks (GANs) have enjoyed much success in learning high-dimensional distributions. Learning objectives approximately minimize an $f$-divergence ($f$-GANs) or an integral probability metric (Wasserstein GANs) between the model and the data distribution using a discriminator. Wasserstein GANs enjoy superior empirical performance, but in $f$-GANs the discriminator can be interpreted as a density ratio estimator which is necessary in some GAN applications. In this paper, we bridge the gap between $f$-GANs and Wasserstein GANs (WGANs). First, we list two constraints over variational $f$-divergence estimation objectives that preserves the optimal solution. Next, we minimize over a Lagrangian relaxation of the constrained objective, and show that it generalizes critic objectives of both $f$-GAN and WGAN. Based on this generalization, we propose a novel practical objective, named KL-Wasserstein GAN (KL-WGAN). We demonstrate empirical success of KL-WGAN on synthetic datasets and real-world image generation benchmarks, and achieve state-of-the-art FID scores on CIFAR10 image generation.

In the current level of evolution of Soccer 3D, motion control is a key factor in team's performance. Recent works takes advantages of model-free approaches based on Machine Learning to exploit robot dynamics in order to obtain faster locomotion skills, achieving running policies and, therefore, opening a new research direction in the Soccer 3D environment. In this work, we present a methodology based on Deep Reinforcement Learning that learns running skills without any prior knowledge, using a neural network whose inputs are related to robot's dynamics. Our results outperformed the previous state-of-the-art sprint velocity reported in Soccer 3D literature by a significant margin. It also demonstrated improvement in sample efficiency, being able to learn how to run in just few hours. We reported our results analyzing the training procedure and also evaluating the policies in terms of speed, reliability and human similarity. Finally, we presented key factors that lead us to improve previous results and shared some ideas for future work.

Despite of the recent progress in agents that learn through interaction, there are several challenges in terms of sample efficiency and generalization across unseen behaviors during training. To mitigate these problems, we propose and apply a first-order Meta-Learning algorithm called Bottom-Up Meta-Policy Search (BUMPS), which works with two-phase optimization procedure: firstly, in a meta-training phase, it distills few expert policies to create a meta-policy capable of generalizing knowledge to unseen tasks during training; secondly, it applies a fast adaptation strategy named Policy Filtering, which evaluates few policies sampled from the meta-policy distribution and selects which best solves the task. We conducted all experiments in the RoboCup 3D Soccer Simulation domain, in the context of kick motion learning. We show that, given our experimental setup, BUMPS works in scenarios where simple multi-task Reinforcement Learning does not. Finally, we performed experiments in a way to evaluate each component of the algorithm.