Wilson Mar bio photo

Wilson Mar

Hello!

Email me Calendar Skype call

LinkedIn Twitter Gitter Instagram Youtube

Github Stackoverflow Pinterest

Brand names for how corporate overlords are making humans into robots

US (English)   Español (Spanish)   Français (French)   Deutsch (German)   Italiano   Português   Cyrillic Russian   中文 (简体) Chinese (Simplified)   日本語 Japanese   한국어 Korean

Overview

Leading AI Companies

In the 201X’s there is an “arms race” in offering Artificial intelligence and Machine Learning (ML) services in their clouds:

Each of the above are cloud vendors may cash in by charging for storing and processing data on their cloud. Facebook, Instagram, and Google, as we all know to our chagrin, makes money from selling their user’s data to advertisers.

In 2016, AI researchers from six major tech companies, including Apple, Amazon, and Google, formed the Partnership on AI (PAI). In 2018, Baidu (the first from China) joined the now 70 member organizations that include Tufts University and Wikimedia.

Alternative clouds

H2O.ai

plainsight.ai focuses on computer vision, using Google CoLab.

floydhub.com (alas, went bankrupt)

Pubs

Benedict Evans, resident futurist at venture capital firm Andreessen Horowitz, observes in a blog post that the future of AI remains opaque: “This field is moving so fast that it’s not easy to say where the strongest leads necessarily are, nor to work out which things will be commodities and which will be strong points of difference.”

awesome-machine-learning provides many links to resources, so they will not be repeated here.

There is a website that specializes in academic publications about Artificial Intelligence. See the Arxiv Paper Analysis Worksheet (Responses) on Google Sheet

Microsoft Academic Graph (MAG) knowledge base mined from the Bing web index. It models scholarly activities: field of study, author, institution, paper, venue, and event.


Algorithmia

Algorithmia.com provide API interfaces to algorithms offered by its partners. They have these data conversion utilities for conventional lookups of data:

  • https://algorithmia.com/algorithms/alixaxel/CoordinatesToTimezone

  • https://algorithmia.com/algorithms/Geo/ZipData

  • https://algorithmia.com/algorithms/Geo/ZipToState

  • https://algorithmia.com/algorithms/Geo/LatLongDistance

  • https://algorithmia.com/algorithms/Geo/LatLongToUTM

  • https://algorithmia.com/algorithms/util/ip2hostname

  • https://algorithmia.com/algorithms/opencv/ChangeImageFormat (from jpg to png)


Benefits from AI/ML in Business

Deloitte’s “State of AI in the Enterprise, 3rd Edition” study

  • Cost reduction: Applying AI and intelligent automation solutions to automate tasks that are relatively low value and often repetitive, reducing costs through improved efficiency and quality.

  • Speed to execution: Reducing the time required to achieve operational and business results by minimizing latency.

  • Reduced complexity: Improving understanding and decision making through analytics that are more proactive, predictive, and able to see patterns in increasingly complex sources.

  • Transformed engagement: Changing the way people interact with technology, enabling businesses to engage with people on human terms rather than forcing humans to interact on machine terms.

  • Fueled innovation: Redefining where to play and how to win by using AI to enable innovative new products, markets and business models.

  • Fortified trust: Securing a business from risks such as fraud and cyber; Improving quality and consistency while enabling greater transparency to enhance brand trust.


Translation

https://translate.google.com and the Google Translate API has been working on translating websites since the 90’s. In 2017 Google made a breakthrough

Microsoft’s Translator Speech

Computer Vision OpenCV

Open-source OpenCV (Computer Vision) was an early entrant and its C code is efficient and has been refined over time. It’s still used today by many because it is free open-sourced as well.

Coursera (DeepLearning.ai) has a 3 COURSE SPECIALIZATION: Generative Adversarial Networks (GANs): 1) Build basic GANs, 2) Build Better GANs, 3) Apply GANs.

Generative Adversarial Networks (GANs) are powerful Machine Learning (ML) models capable of generating realistic image, video, and voice outputs.

  • http://www.michaelhasey.com/gan-exterior
  • https://www.medium.com/syncedreview/gan-2-(3439d33ebaf

“Adversarial” because GAN techniques are rooted in game theory. Its use cases include:

  • improving cybersecurity by fighting against adversarial attacks and
  • anonymizing data to preserve privacy to generating state-of-the-art images,
  • colorizing black and white images,
  • increasing image resolution,
  • creating avatars,
  • turning 2D images to 3D
  • etc.

Microsoft’s Computer Vision

https://docs.microsoft.com/en-us/azure/cognitive-services/custom-vision-service/getting-started-build-a-classifier Hands-on guide: build a classifier with Custom Vision

Microsoft’s “Face”

  • https://algorithmia.com/algorithms/z/ColorPalettefromImage

  • https://algorithmia.com/algorithms/opencv/FaceDetection then https://algorithmia.com/algorithms/opencv/CensorFace

  • https://algorithmia.com/algorithms/ocr/RecognizeCharacters OCR

Some of these make use of OpenCV (CV = Computer Vision).

Google Cloud Vision API

Voice Recognition

Microsoft’s Web App Bot

NLP Sentiment Analysis

Analyze text for positive or negative sentiment (opinion), based on a training database of potential word meanings, which involved Natural Language Processing:

  • https://algorithmia.com/algorithms/nlp/SentimentAnalysis

  • IBM’s algorithm

Andrew W. Trask, PhD student at University of Oxford Deep Learning for Natural Language Processing authored Grokking Deep Learning.

Use Bag of words and Word2vec transform words into vectors. Use TFLearn, a Python library for quickly building networks.

Document (article) Search

Google made it’s fortune on offering search services.

Microsoft’s Bing Search

TF-IDF = Term Frequency - Inverse Document Frequency emphasizes important words (called a vector) which appear rarely in the corpus searched (rare globally). which appear frequently in document (common locally) Term frequency is measured by word count (how many occurances of each word).

The IDF to downweight words is the log of #docs divided by 1 + #docs using given word.

Cosine similarity normalizes vectors so small angle thetas identify similarity.

Normalizing makes the comparison invariant to the number of words. The common compromise is to cap maximum word count.

Recommender

Recommender systems recommend (advises) users about what to do, based on the pattern detected in similar situations observed in the past.

collaborative filtering and factorization machines.

implement the solution using sparse distributed matrices in PySpark.

Footnotes

https://www.wikiwand.com/en/Deep_learning

Social communities

More

This is one of a series on AI, Machine Learning, Deep Learning, Robotics, and Analytics:

  1. AI Ecosystem
  2. Machine Learning
  3. Testing AI

  4. Microsoft’s AI
  5. Microsoft’s Azure Machine Learning Algorithms
  6. Microsoft’s Azure Machine Learning tutorial
  7. Microsoft’s Azure Machine Learning certification

  8. Python installation
  9. Juypter notebooks processing Python for humans

  10. Image Processing
  11. Tessaract OCR using OpenCV
  12. Amazon Lex text to speech

  13. Code Generation

  14. Multiple Regression calculation and visualization using Excel and Machine Learning
  15. Tableau Data Visualization