Automated Machine Learning & Dialog Systems

Notes:

Automated Machine Learning (AutoML) refers to the use of machine learning algorithms and techniques to automatically select and optimize machine learning models for a given task, without the need for manual intervention or expertise in machine learning. AutoML can be used to automate many aspects of the machine learning process, including data preprocessing, feature engineering, model selection, and hyperparameter optimization.

In the context of dialog systems, AutoML can be used to automate the process of building and optimizing natural language processing (NLP) models that are used to understand and generate human-like responses in a conversation. For example, an AutoML system might be used to automatically select the best machine learning model for a task such as language translation or text classification, based on a set of input data and a specified performance metric.

AutoML can be particularly useful in dialog systems when the data used to train the NLP models is large and varied, or when the performance requirements are high and require frequent model updates. By automating the process of model selection and optimization, AutoML can help developers and researchers build more effective and efficient dialog systems faster and with less effort.

Resources:

auto-sklearn .. automated machine learning toolkit (drop-in replacement for a scikit-learn estimator)
autograd .. efficiently computes derivatives of numpy code
autokit .. automatic machine learning suite
autoweka .. selects a learning algorithm and sets its hyperparameters
hyperopt .. distributed asynchronous hyperparameter optimization in python

Wikipedia:

References:

Data and artificial intelligence strategy: A conceptual enterprise big data cloud architecture to enable market-oriented organisations
C Moreno, RAC González, EH Viedma – IJIMAI, 2019 – dialnet.unirioja.es
… will allow increasing their power and the industrialization of them: Cloud AI and AutoML (Automated machine learning) … Automatic Machine Learning (AutoML) is a software that can run in the cloud or in a … the shortage of specialized professional [27], the use of AutoML tools [28 …

A blockchain and automl approach for open and automated customer service
Z Li, H Guo, WM Wang, Y Guan… – IEEE Transactions …, 2019 – ieeexplore.ieee.org
… In or- der to overcome these challenges, an open and automated customer service platform based on Internet of things (IoT), blockchain, and automated machine learning (AutoML) is proposed. The data are gathered with the use of IoT de- vices during the customer service …

Object Recognition with Machine Learning: Case Study of Demand-Responsive Service
PJ Lin, S Hung, SFS Lam… – 2019 IEEE International …, 2019 – ieeexplore.ieee.org
… combination with the AutoML Vision API. A chatbot is an application that simulates human conversation or chat and attempts to build a program in which people believe that they are chatting with other humans. In recent years, chatbots have been seen in social websites …

A Smart Personal AI Assistant for Visually Impaired People: A Survey
G Singh, K Takhtani, O Kandale, N Dadhwal – 2019 – academia.edu
… user using voice command, recognizing and analyzing the surrounding environment and providing suitable response, chat bot … Along with that a chatbot that will answer user’s query is implemented … It uses AutoML Vision Edge to build and deploy fast, high-accuracy models to …

Conditional Parallel Coordinates
DKI Weidele – 2019 IEEE Visualization Conference (VIS), 2019 – ieeexplore.ieee.org
… in AutoML Parallel Coordinates have recently been applied in visualizations of automatic machine learning software [15 … Samu- lowitz and Dakuo Wang for the provision of the AutoML data set … Atmseer: Increasing transparency and controllability in automated machine learning …

Implementing a college enquiry chatbot
UK Bavishi – 2019 – dspace.calstate.edu
… All these advantages of a chatbot constitute the motivation to implement a College Enquiry Chatbot. Before implementing College Enquiry Chatbot, various existing chatbots were reviewed such as Amazon Shopping App [3], Alexa [4], Bank of America (Erica bot) [5] Page 16. 3 …

NeuronBlocks: Building Your NLP DNN Models Like Playing Lego
M Gong, L Shou, W Lin, Z Sang, Q Yan, Z Yang… – arXiv preprint arXiv …, 2019 – arxiv.org
… Sequence generation task. Sequence genera- tion is widely used in NLP fields such as ma- chine translation (Bahdanau et al., 2015), text summarization (See et al., 2017), and dialogue systems (Wen et al., 2015). • AutoML (Elsken et al., 2019) …

A Human-in-the-Loop Method for Developing Machine Learning Applications
L Yang, M Li, J Ren, C Zuo, J Ma… – 2019 6th International …, 2019 – ieeexplore.ieee.org
… is mainly divided into two parts: 1) The automated machine learning development method … data processing loop based on the negative feedbacks: When AutoML is used … NLP tasks, relation extraction, event extraction, knowledge graph, machine translation and dialogue system …

Implementation of AI
T Taulli – Artificial Intelligence Basics, 2019 – Springer
… For example, if you are developing a custom chatbot for customer service, you might want to measure against metrics like the resolution rate and … During the past few years, there has also emerged a new category of AI Tools called automated machine learning or autoML …

Salus: Fine-Grained GPU Sharing Primitives for Deep Learning Applications
P Yu, M Chowdhury – arXiv preprint arXiv:1902.04610, 2019 – arxiv.org
… in re- cent years across many data-driven application domains, ranging from machine translation and image captioning to chat bots and personal … Additionally, the increasingly popular trend of automatic hyper-parameter tuning of DL models (eg, AutoML [18,36, 42]) further …

Machine Learning with Decision Trees and Multi-Armed Bandits: An Interactive Vehicle Recommender System
T Yu, O Mengshoel, D Meroux, Z Jiang – 2019 – sae.org
… Customers can then provide information to and interact with a chatbot on the website to get suitable recommendations … applied in different areas, such as online recommender systems [12,13], online advertising [12], chatbots [19], and automated machine learning [21,22] …

Effective Use of the Machine Learning Approaches on Different Clouds
Y Kumar, S Kaul, K Sood – Available at SSRN 3355203, 2019 – papers.ssrn.com
… A session of question answers from users and chatbots can also take place and for … The main intention of using these chatbot as humans is to make users feel its … and version tools • Easy data contact • Modeling API’s Microsoft Azure • Automated machine learning • Support for …

Design feasibility of an automated, machine-learning based feedback system for motivational interviewing.
ZE Imel, BT Pace, CS Soma, M Tanana, T Hirsch… – …, 2019 – psycnet.apa.org
… Design feasibility of an automated, machine-learning based feedback system for motivational interviewing. Psychotherapy, 56(2), 318-328 … For example, a text-based chatbot client could interact with a trainee therapist and provide instant feedback on every therapist statement …

A survey on practical applications of multi-armed and contextual bandits
D Bouneffouf, I Rish – arXiv preprint arXiv:1904.10040, 2019 – arxiv.org
… paradigm of contextual bandits as frame- work for proactive dialog systems … Finance ? Dynamic pricing ? Recommendr system ? ? ? ? Maximization ? Dialogue system ? Telecomunication ? Anomaly … be further used as a tool for advancing automated machine learning …

Distributed Deep Learning for Broadly Deployed Health Care
BC Ooi, G Chen, Q Lin, KY Ngiam, S Wang, W Wang… – pdfs.semanticscholar.org
… Abstract—While chatbots and self-driving cars may be the most prominent exemplars of deep learning systems deployed in society, personalized health care is perhaps … To ease the burden on AI users, the concept of AutoML (automated machine learning) has been suggested …

Data-and Expert-Driven Analysis of Cause-Effect Relationships in the Production of Lithium-Ion Batteries
T Komas, R Daub, MZ Karamat… – 2019 IEEE 15th …, 2019 – ieeexplore.ieee.org
… The ASSS is based on humanY computer interaction and automated machine learning (autoML), which enables … m. blum, and F. hutter, Iefficient and robust automated machine Learning,” in advances … Ihow software Developers mitigate collaboration Friction with chatbots,” Feb …

An Approach to Deep Learning Service Provision with Elastic Remote Interfaces
M Sun, Z Yang, H Wu, Q Liu, X Liu – International Conference on Artificial …, 2019 – Springer
… range of deep learning product services, such as Amazon Polly [9] for text-to-speech, Microsoft Azure’s Emotion API can be used to analyze human emotions, and the Recognition API verifies speaker recognition [10], Google’s Dialogflow chat bot can be … AutoML Vision … Chatbot …

Data scientists
T Stadelmann, K Stockinger, GH Bürki… – Applied Data Science, 2019 – Springer
… 11See http://xkcd.com/1425/. While the described phenomenon might be easy to solve in the year 2018, a contemporary example would be chatbots. 12The same applies to automated machine learning, although such systems have a real value for certain applications …

Meta Learning for Hyperparameter Optimization in Dialogue System
JT Chien, WX Lieow – Proc. Interspeech 2019, 2019 – isca-speech.org
… a multifidelity Bayesian optimization for hyperparameter optimization in reinforcement learning for dialogue system … in deep reinforcement learning for task-oriented dialogue systems,” in Prof … based bandit optimization with ap- plication to automatic machine learning,” in Artificial …

A survey on bias and fairness in machine learning
N Mehrabi, F Morstatter, N Saxena, K Lerman… – arXiv preprint arXiv …, 2019 – arxiv.org
… systems affect our daily lives with their inherent biases, such as the existence of bias in AI chatbots, employment matching … compared SAVRY, a tool used in risk assessment frameworks that includes human intervention in its process, with automatic machine learning methods in …

AIR5: Five Pillars of Artificial Intelligence Research
YS Ong, A Gupta – IEEE Transactions on Emerging Topics in …, 2019 – ieeexplore.ieee.org
… The overall approach falls under the scope of so-called AutoML (automated machine learning [39]), a topic … At the leading edge of AutoML is an ongoing attempt to develop algorithms that … even providing psychological support to Syrian refugees through the use of chat-bots [54] …

Automatically responding to customers
R Huijzer – pure.tue.nl
… The IBM sales department claims that Autodesk using chatbots cut down their resolution … Conversational agent or dialogue systems aim to communicate with humans using natural lan- guage. Consensus is not clear on whether a chatbot is synonymous to conversational agent …

Chameleon: A Language Model Adaptation Toolkit for Automatic Speech Recognition of Conversational Speech
Y Song, D Jiang, W Zhao, Q Xu, RCW Wong… – Proceedings of the 2019 …, 2019 – aclweb.org
… In addition, the hyperparameter tun- ing step can be combined with current Automatic Machine Learning (AutoML) techniques (Quan- ming et al., 2018) to achieve better performance and user experience … Scalable language model adaptation for spo- ken dialogue systems …

The era of connectedness: How AI will help deliver the future of banking
T Lau, B Leimer – Journal of Digital Banking, 2019 – ingentaconnect.com
… officer (CEO) Satya Nadella recently showed off Xiaoice, a social chatbot, which can … The power of AI and automated machine learning is already being felt in daily life … AI is also prevalent in navigation applications, streaming services, virtual assistants, chatbots, ride-sharing and …

Machine Learning with Microsoft Technologies
L Etaati – 2019 – Springer
… 93 Decision Trees….. 100 Automated Machine Learning Inside Power Query … 335 Question-and-Answer Chatbot …

Conversational agents and negative lessons from behaviourism
M Gnjatovi? – Innovations in Big Data Mining and Embedded …, 2019 – Springer
… One of the most obvious advantages of automated machine learning analyses of large-scale dialogue corpora … to implement socially believable 2 conversational agents whose functionalities go substantially beyond chatbots … Jokinen, K., McTear, M.: Spoken Dialogue Systems …

Beyond validation accuracy: incorporating out-of-distribution checks, explainability, and adversarial attacks into classifier design
JS Hyatt, MS Lee – … Intelligence and Machine Learning for Multi …, 2019 – spiedigitallibrary.org
SPIE Digital Library Proceedings.

Artificial intelligence and the technologies of the Next Production Revolution
A Nolan – 2019 – oecd-ilibrary.org
… Furthermore, the use of AI in production will be spurred by automated machine learning processes that can help businesses, scientists and other … Chatbots and similar systems are now able to interact with workers using natural language, providing answers and context-specific …

Engineering Reliable Deep Learning Systems
P Santhanam, E Farchi, V Pankratius – arXiv preprint arXiv:1910.12582, 2019 – arxiv.org
… mistakes can have a more severe social consequence: In 2016, Microsoft launched an experimental AI chat bot, called Tay … Symmetry Networks”, Advances in Neural Information Processing Systems 27 (2014) [32] IBM Research Blog on “Automated Machine Learning and …

Raising Awareness of Artificial Intelligence for Transportation Systems Management and Operations
D Gettman – 2019 – rosap.ntl.bts.gov
… Interoperability among neural network modeling systems and frameworks via Open Neural Network Exchange (ONNX).3 • Automated machine learning with AutoML—speeding the process of … It is generally accepted that no chatbot is yet “thinking” or represents “strong” AI …

The ICT Tsunami and Your Future
S Birudavolu, B Nag – Business Innovation and ICT Strategies, 2019 – Springer
… By October 2017, Google announced that the AutoML software was building better AI software than the human engineers themselves, with regard to … Banking: Connecting to a million customers today may be achieved with an intelligent chatbot, such as State Bank of India’s …

Current Developments On Artificial Intelligence And Liability For Robot Caused Damages
S Yünlü – YED?TEPE ÜN?VERS?TES? HUKUK FAKÜLTES? … – law.yeditepe.edu.tr
… In this regard, the bots referred to as “chatbots” must be mentioned … com/technology/2016/jun/ 28/chatbot-ai-lawyer-donotpay- parking-tickets-london-new-york, Accessed 20.03 … co. uk/life-style/gadgets-and-tech/news/google-child-ai-bot- nasnet-automl-machine-learning-artificial …

Individualized Treatment Strategy for Depressive Disorder
J Chen, S Hu – Depressive Disorders: Mechanisms, Measurement and …, 2019 – Springer
… An automated machine-learning speech classifier was performed in predicting psychosis onset and reached an accuracy of 79% (Corcoran et al … the availability and effectiveness of existing treatments, such as Internet-based Cognitive Behavioral Therapy(CBT), CBT chatbots …

Keeping Your AI Under Control
A Tamboli – Springer
… “In today’s world of cloud services and AUTOML [automated machine learning] frameworks, the … language processing in the form of compilers, this appli- cation (online chatbot) still struggles a lot … out of the customer service sector and replacing them with chatbots or automated …

Artificial Intelligence: Can Information be Transformed into Intelligence in Surgical Education?
AY Sheikh, JI Fann – Thoracic Surgery Clinics, 2019 – thoracic.theclinics.com
Skip to Main Content …

The AIGO
J Seger, N Miailhe, S Mueller – 2019 – thefuturesociety.org
… US Citizenship and Immigration Services (USCIS) launched a chatbot named “Emma … The respective provider’s data centers handle the actual computation; results obtained can be can transferred to internal IGO IT infrastructure through APIs.53 AutoML, the newest innovation in …

Artificial intelligence and data analytics: Emerging opportunities and challenges in financial services
C Coombs, R Chopra – 2019 – repository.lboro.ac.uk
… In the front office, AI can be used in the form of chatbots to speed- up and streamline the claims process for consumers and mitigate the number and magnitude of fraudulent claims for the insurers … Google has developed AutoML for this purpose …

The Technology Adoption and Governance of Artificial Intelligence in the Philippines
RS Concepcion, RAR Bedruz, AB Culaba… – 2019 IEEE 11th … – ieeexplore.ieee.org
… Automated machine learning (AML) through AI will transform business analytics so business analysts and experts could focus … birth to another AI on its own in the face of AutoML for NASNet … But the impact of AI automation will vanish majority of the employees by using chat bots …

Requirements for Developing Robust Neural Networks
JS Hyatt, MS Lee – arXiv preprint arXiv:1910.02125, 2019 – arxiv.org
… Fusi, N.; Sheth, R.; and Elibol, MH 2018. Probabilistic Matrix Factorization for Automated Machine Learning. In Advances in Neural Information Processing Systems 31, 3348–3357 … Hunt, E. 2016. Tay, Microsoft’s AI Chatbot, Gets a Crash Course in Racism From Twitter …

Engineering Applications of Neural Networks: 20th International Conference, EANN 2019, Xersonisos, Crete, Greece, May 24-26, 2019, Proceedings
J Macintyre, L Iliadis, I Maglogiannis, C Jayne – 2019 – books.google.com
… Health Records: From temporal Abstraction to Timeseries Interpretability” A three-hour tutorial on “Automated Machine Learning for Bioinformatics … for the disabled; smart/assistive environments; sensor networks for smart environ- ments; dialogue systems; telemedicine; virtual …

High-performance medicine: the convergence of human and artificial intelligence
EJ Topol – Nature medicine, 2019 – nature.com
The use of artificial intelligence, and the deep-learning subtype in particular, has been enabled by the use of labeled big data, along with markedly enhanced computing power and cloud storage, across all sectors. In medicine, this is beginning to have an impact at three levels: for …

Developing Cognitive Bots Using the IBM Watson Engine
W Platform – Springer
… She has implemented and worked on chat bots using various cognitive engines, including Watson, LUIS … scale due to the advent of cognitive virtual assistants—more commonly known as chatbots. A chatbot is a computer program that converses in natural language via auditory …

An automated machine learning based decision support system to predict hotel booking cancellations
N Antonio, A de Almeida… – An automated machine …, 2019 – repositorio.iscte-iul.pt
… 2019. An Automated Machine Learning Based Decision Support System to Predict Hotel Booking Cancellations. Data Science Journal, 18: 32, pp. 1–20 … RESEARCH PAPER An Automated Machine Learning Based Decision Support System to Predict Hotel Booking …

Learning navigation behaviors end-to-end with autorl
HTL Chiang, A Faust, M Fiser… – IEEE Robotics and …, 2019 – ieeexplore.ieee.org
… Hyper-parameter optimization in deep learning: AutoML methods learn the neural network architecture for a given problem. AutoML achieves this with gradient descent [2], reinforcement learning [38], and even hyper-parameter opti- mization [34] …

Beauty Is in the AI of the Beholder: Artificial and Superintelligence
N Lee – The Transhumanism Handbook, 2019 – Springer
… In July 2017, AI chat bots at Facebook invented its own languages for chattering and better at negotiation [52] (better negotiating for world … In December 2017, Google announced that AutoML has beaten the human AI engineers at their own game by building machine-learning …

Generalizing from a few examples: A survey on few-shot learning
Y Wang, Q Yao, J Kwok, LM Ni – arXiv: 1904.05046, 2019 – researchgate.net
Page 1. 1 Generalizing from a Few Examples: A Survey on Few-Shot Learning YAQING WANG, Hong Kong University of Science and Technology and Baidu Research QUANMING YAO?, 4Paradigm Inc. JAMES T. KWOK, Hong …

Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics
A Korhonen, D Traum, L Màrquez – … of the 57th Annual Meeting of the …, 2019 – aclweb.org
Page 1. ACL 2019 The 57th Annual Meeting of the Association for Computational Linguistics Proceedings of the Conference July 28 – August 2, 2019 Florence, Italy Page 2. Diamond Sponsors: Platinum Sponsors: ii Page 3. Gold sponsors: Keiosk Analytics Silver sponsors …

Toward AI Systems that Augment and Empower Humans by Understanding Us, our Society and the World Around Us
J Crowley, AP O’Sullivan, A Nowak, C Jonker… – 2019 – humane-ai.eu
… Compositionality and automated machine learning (Auto-ML) … The recent success of general-purpose algorithm configuration and selection methods—and notably, the rise of automated machine learning already leverages this insight …

Legal Issues And Computational Measures At The Cross-Section Of Ai, Law And Policy
S Rathi – 2019 – web2py.iiit.ac.in
… 42 6.3.3 AutoML … It is important that autonomous systems remain value neutral. • Interactions: With the growing penetration of chat bots and other interactive forms of AI, it is a value add to these technologies if they can indulge in coherent and consistent interactions …

Intelligent Process Automation: Building the bridge between Robotic Process Automation and Artificial Intelligence
M Bellman, G Göransson – 2019 – diva-portal.org
Page 1. IN DEGREE PROJECT INDUSTRIAL ENGINEERING AND MANAGEMENT, SECOND CYCLE, 30 CREDITS , STOCKHOLM SWEDEN 2019 Intelligent Process Automation Building the bridge between Robotic Process Automation and Artificial Intelligence …

Audiovisual Algorithms
A Schweikert – 2019 – nyu.edu
… On the other 23 hand, automated reasoning, which can be used to power chatbots or other artificial interactions, is … such as playlists that recommend new music based on your listening habits and chatbots that attempt to provide online customer service …

On The Danger of Artificial Intelligence
S Samiei – 2019 – openrepository.aut.ac.nz
Page 1. 1 | Page On The Danger of Artificial Intelligence Saba Samiei A thesis submitted to Auckland University of Technology in fulfilment of the requirements for the degree of Master of Computing and Information Sciences (MCIS) July 2019 …

Judging machines: philosophical aspects of deep learning
A Schubbach – Synthese, 2019 – Springer
Although machine learning has been successful in recent years and is increasingly being deployed in the sciences, enterprises or administrations, it has ra.

A strategic perspective on the commercialization of artificial intelligence: a socio-technical analysis
S Ray Barua – 2019 – dspace.mit.edu
Page 1. A Strategic Perspective on the Commercialization of Artificial Intelligence: A socio-technical analysis by Siddhartha Ray Barua BE (Information Science & Engineering), MSRIT (2007) PGDIM (Industrial Management), NITIE (2012) …

The Disruptions of 5G on Data-driven Technologies and Applications
D Loghin, S Cai, G Chen, TTA Dinh, F Fan… – arXiv preprint arXiv …, 2019 – arxiv.org
Page 1. arXiv:1909.08096v1 [cs.NI] 6 Sep 2019 1 The Disruptions of 5G on Data-driven Technologies and Applications Dumitrel Loghin, Member, IEEE, Shaofeng Cai, Gang Chen, Member, IEEE, Tien Tuan Anh Dinh, Feiyi Fan …

FinTech Challengers and Incumbents’ Responses: A Window into Innovation Strategy Modes within Australia’s Financial Services Sector
M Gold, P Ali – Available at SSRN 3467855, 2019 – papers.ssrn.com
… programming interfaces or ‘APIs’) which improve customer experience (CX) or user experience (UX), especially important when provisioning standardised products via ubiquitous devices and digital distribution platforms; data analytics and automated machine-learning tools for …

A Case for a New IT Ecosystem: On-The-Fly Computing
H Karl, D Kundisch, FM auf der Heide… – Business & Information …, 2019 – Springer
… As a consequence, neither template nor free techniques are applied on a wide scale today. An exception to this are domain-specific configurations, an example being automatic assembly of machine-learning applications (AutoML, eg, Thornton et al. 2013) …

A Global Fintech Overview
GY Allayannis, JM Becker – Darden Case No. UVA-F-1860, 2019 – papers.ssrn.com
Page 1. UVA-F-1860 Apr. 26, 2019 This technical note was prepared by Joseph M. Becker (MBA ’19) and George (Yiorgos) Allayannis, Paul Tudor Jones II Professor of Business Administration. Copyright © 2019 by the University …

An overview of the NFAIS Conference: Artificial intelligence: Finding its place in research, discovery, and scholarly publishing
B Lawlor – Information Services & Use, 2019 – content.iospress.com
Page 1. uncorrected proof version Information Services & Use 1 (2019) 1–32 1 DOI 10.3233/ISU-190068 IOS Press An overview of the NFAIS Conference: Artificial intelligence: Finding its place in research, discovery, and scholarly publishing …

From user feedback to requirements using chatbots
D Horváth – 2019 – dspace.library.uu.nl
… Furthermore, chatbots can help in guiding the users through the bug reporting process by controlling the flow of the conversation and using follow-up … After the evaluation the subjects were asked to rate the chatbot using a usab- ility scale, and were interviewed to find out their …

Designing and Building Evidence-Based Intelligent Algorithmic Systems in Online Communities
B Yu – 2019 – conservancy.umn.edu
… and spelling issues on Stack Overflow [17] to improve community content quality; Kerly et al. developed chatbots into education domains for Intelligent Tutoring [48]. In Wikipedia, various sophisticated tools are used to work on tasks automatically, such as …

Air learning: An ai research platform for algorithm-hardware benchmarking of autonomous aerial robots
S Krishnan, B Borojerdian, W Fu, A Faust… – arXiv preprint arXiv …, 2019 – arxiv.org
Page 1. Air Learning: An AI Research Platform for Algorithm-Hardware Benchmarking of Autonomous Aerial Robots Srivatsan Krishnan†, Behzad Boroujerdian‡, William Fu†, Aleksandra Faust?, and Vijay Janapa Reddi†‡ † Harvard …

Beyond Labels and Captions: Contextualizing Grounded Semantics for Explainable Visual Interpretation
SN Aakur – 2019 – search.proquest.com
Beyond Labels and Captions: Contextualizing Grounded Semantics for Explainable Visual Interpretation. Abstract. One of the long-standing problems in artificial intelligence is the development of intelligent agents with complete visual understanding …