Big Data Applications & Analytics �Motivation/Overview�Machine Deep Learning, Big Data and the Cloud

Geoffrey Fox

School of Informatics, Computing, and Engineering

Department of Intelligent Systems Engineering

​

​

​

Digital Science Center

Indiana University Bloomington

gcf@indiana.edu

http://www.infomall.org

​

​

1

Deep

Digital Science Center

Digital Science Center

Topics in the course

• A set of modules covering Big Data in different areas with algorithms and applications
• Updates reflect major shift to Deep Learning and stabilizing of clouds�
• Physics: study of Higgs Boson roughly unchanged
• Big Data in Sports: still accurate
• Big Data Use Case Survey; still accurate from application point of view but rather old
• Sensors and Edge Systems; could be improved but general discussion OK and won’t emphasize
• Health Informatics; actively updated
• Commerce; updated includes a little on recommender engines
• Mobility and Transportation Systems: added from Spring class �
• Cloud Computing; mature and updated
• Technologies; used as necessary but deep learning replaces many from older versions of course and are described in methods oriented discussions
• Image-based applications: in transportation and elsewhere
• Times series-based applications: in medicine and environment

​

​

​

​

​

​

2

Digital Science Center

Organizational Details

• I423 E434 and E534 are cross-listed. A detailed listing for the previous version of this course can be found at http://dsc.soic.indiana.edu/publications/E534-BigDataSystems-DeepLearning.pdf
• There are support modules
• Python: https://cybertraining-dsc.github.io/courses/python/  
• Linux: https://cybertraining-dsc.github.io/courses/linux/ (not needed if use Jupyter notebooks)
• Writing in markdown: https://cybertraining-dsc.github.io/courses/markdown/ (text format in GitHub and Jupyter)
• The course is graded by a final project plus homework. The project can be software or essay style but graduate sections need a software project for highest grades. The sections share lecture material but differ in homework and projects.
• Piazza is used for communication: https://piazza.com/iu/fall2020/e534fa20/resources will link to material
• The course focuses on applications using today's technology model of an intelligent cloud linked to an intelligent edge. Upgrades in this year's version include a major module on applications of deep learning reflecting the dominance of deep learning over older statistics and machine learning approaches.
• Any software will be based on Python.

3

Digital Science Center

Using Jupyter Notebook

• Example at https://colab.research.google.com/drive/1XEMvpnQ4uw8F_lOBFi9dWq02gxeVmxIR?usp=sharing
• Notebooks and data stored on Google drive
• Basic version of Colab free -- restrictions on GPU type and usage
• Colab Pro $9.99 a month with better GPU and less restrictions
• Note as a notebook have code, documentation and output in one file
• Can access Github

4

Digital Science Center

Impact of Big Data

• There is an endlessly growing amount of data as we record every transaction between people and the environment (whether shopping or on a social networking site) while smart phones, smart homes, ubiquitous cities, smart power grids, and intelligent vehicles deploy sensors recording even more.
• Science with satellites and accelerators is giving data on transactions of particles and photons at the microscopic scale.
• This data are and will be stored in immense  clouds with co-located storage and computing that perform "analytics" that transform data into information and then to wisdom and decisions; data mining finds the proverbial knowledge diamonds in the data rough.
• This disruptive transformation is driving the economy and creating millions of jobs in the emerging area of "data science".
• We discuss this revolution and its implications for industry education and research
• You need to be Digitally/Data aware, Digitally/Data flexible and have Digital/Data intuition
• The Digital/Data approach is changing everything ….

5

Digital Science Center

Some Trends from 2014

• The Data Deluge is clear trend from Commercial (Amazon, e-commerce) , Community (Facebook, Search) and Scientific applications
• Smaller (INTEL/ARM/AMD) chips drive
• Multicore (i.e. more computing) on shared servers
• Smaller Light weight clients from smartphones, tablets to sensors (i.e. more clients)
• Clouds with cheaper, greener, easier to use IT for applications
• New jobs associated with new curricula
• Clouds and distributed systems
• Data Science and Data Engineering (equally important)
• Deep Learning often best approach replacing older machine learning
• Using AI-specific chips such as GPU’s

​

6

Added Deep Learning and Data Engineering recently, Could add Edge

Digital Science Center

Dominance of Cloud Computing

• 94 percent of workloads and compute instances will be processed by cloud data centers (22% CAGR) by 2021-- only six percent will be processed by traditional data centers (-5% CAGR).
• Hyperscale data centers will grow from 338 in number at the end of 2016 to 628 by 2021. They will represent 53 percent of all installed data center servers by 2021. They form a distributed Compute (on data) grid with some 50 million servers
• Analysis from CISCO Global Cloud Index 2018 https://newsroom.cisco.com/press-release-content?type=webcontent&articleId=1908858

7

7

Number of instances per server

Number of Cloud Data Centers

Number of Public or Private Cloud Data Center Instances

Digital Science Center

�Gartner’s view is perhaps worrisome

8

Digital Science Center

Digital Science Center

Hype Cycle position implies actions

9

Digital Science Center

Gartner: Hype Cycle for Emerging Technologies, 2019

• Published 6 August 2019 - ID G00370466
• Analysts: Brian Burke, David Smith

​

10

Digital Science Center

Gartner: Priority Matrix for Emerging Technologies, 2019

• Published 6 August 2019 - ID G00370466
• Analysts: Brian Burke, David Smith

​

11

Digital Science Center

Hype Cycles and Priority Matrices

• There are over 100 Hype cycles which have increased over last few years so people must like them!
• They are accompanied by a Priority matrix which explains timing and importance of each innovation.
• Figure maps Priority matrix cells to investment approaches

12

Digital Science Center

Gartner: Hype Cycle for Emerging Technologies, 2008

• Hype Cycle for Emerging Technologies, 2008
• ARCHIVED Published: 09 July 2008 ID: G00159496

13

Digital Science Center

Gartner: Priority Matrix for Emerging Technologies, 2008

• Hype Cycle for Emerging Technologies, 2008
• ARCHIVED Published: 09 July 2008 ID: G00159496

14

Digital Science Center

15

Gartner: Hype Cycle for Emerging Technologies, 2014

Digital Science Center

Gartner: How Emerging Technology Trends Move Along the Hype Cycle�

• Hype Cycle for Emerging Technologies, 2017
• Published: 21 July 2017 ID: G00314560
• Analyst(s): Mike J. Walker
• The emerging technologies on the 2017 Hype Cycle reveal three distinct megatrends

16

Digital Science Center

Gartner 2020

17

No Clouds

No Quantum Computing

Digital Science Center

Digital Science Center

Gartner: Hype Cycle for Artificial Intelligence, 2019

• Published: 25 July 2019 ID: G00369840
• Analysts: Svetlana Sicular, Jim Hare, Kenneth Brant

18

Digital Science Center

Gartner: Priority Matrix for Artificial Intelligence, 2019

• Published: 25 July 2019 ID: G00369840
• Analysts: Svetlana Sicular, Jim Hare, Kenneth Brant

19

Digital Science Center

Gartner: Hype Cycle for Cloud Computing, 2019

• Published: 8 August 2019 ID: G00370239
• Analysts: David Smith, Ed Anderson, Leah Ciavardini

20

Digital Science Center

Gartner: Priority Matrix for Cloud Computing, 2019

• Published: 8 August 2019 ID: G00370239
• Analysts: David Smith, Ed Anderson, Leah Ciavardini

​

21

Digital Science Center

Hype Cycles

22

Digital Science Center

Digital Science Center

�Big Data Trends

23

Digital Science Center

Digital Science Center

24

http://cs.metrostate.edu/~sbd/ �Oracle ~2010

Digital Science Center

Course Motivation

25

Digital Science Center

26

Digital Science Center

27

Digital Science Center

28

Digital Science Center

29

9/7/2019

Digital Science Center

30

9/7/2019

Digital Science Center

31

Ruh VP Software GE in 2012 http://fisheritcenter.haas.berkeley.edu/Big_Data/index.html

Digital Science Center

32

Ruh VP Software GE http://fisheritcenter.haas.berkeley.edu/Big_Data/index.html

Digital Science Center

33

Ruh VP Software GE http://fisheritcenter.haas.berkeley.edu/Big_Data/index.html

MM = Million

Digital Science Center

�Computing Trends

34

Digital Science Center

Digital Science Center

Lots of Computers are needed

• Performance of individual CPU’s is limited
• Use lots of cores per chip
• Use lots of chips per job
• Superman versus a bunch of people

35

Digital Science Center

36

Digital Science Center

37

9/7/2019

https://en.wikipedia.org/wiki/FLOPS

Digital Science Center

38

9/7/2019

Digital Science Center

39

Digital Science Center

40

Digital Science Center

Overall Global AI and Modeling Supercomputer GAIMSC

41

By Donald Kossmann, Microsoft

Digital Science Center

Issues with a US AI Supercomputer

INTEL REPORTEDLY HOLDS UP US GOVT'S SUPERCOMPUTER LAUNCH, WHICH WILL LIKELY SPUR INCREASED TECH INVESTMENT FOR NATIONAL DEFENSE: Intel's 7-nanometer (7nm) production delays are reportedly interfering with the company's ability to deliver experimental chips for what could have been the world's most advanced supercomputer, according to a report from The New York Times. In 2019, the US Energy Department selected Intel to provide special-made chips for Aurora (at Argonne National Lab), a $500 million supercomputer that promised to deliver AI breakthroughs in climate modeling, medical research, and nuclear simulations. Aurora had the potential to be the world's first computer to reach the exascale computational threshold, but with the delays, China will likely leapfrog the US, considering its three concurrent efforts to develop exascale supercomputers.

The Aurora holdup alludes to the larger phenomenon of the US semiconductor manufacturing industry lagging behind global competitors. Its Q2 2020 earnings report from July, Intel announced that it was running 12 months behind internal targets for developing 7nm manufacturing capabilities. This represented a significant setback both for Intel — which subsequently lost around $50 billion in market cap, per the Times — and the US government. For national security purposes, the US government has attempted to bolster domestic chipmaking capabilities, and Intel is one of the few US semiconductor companies that still develops chips in-house. Other leading US companies such as Apple, Nvidia, and AMD outsource chip manufacturing to either South Korea-based Samsung or Taiwan-based TSMC, which operate the world's most advanced foundries. In fact, while Intel is struggling to manufacture chips on the 7nm standard, TSMC announced just this week that it was on track to begin full volume production on the 3nm standard, which is a full two generations ahead.

​

42

Digital Science Center

As Intel has problems, NVIDIA succeeds

Nvidia's business transformation embracing cloud and AI services has been a resounding success, with Data Centers revenues growing 167% year-over-year (YoY) for the company's fiscal Q2 2021 (ended July 26, 2020). In that same quarter, Nvidia's revenue from Data Centers surpassed that from Gaming, making it the company's highest-grossing business segment. Nvidia specializes in developing graphics processing units (GPUs), which were originally used to render graphics for video games. For example, three years after its founding in 1997, Nvidia landed a major deal to supply Microsoft with graphics chips for the original Xbox console, according to Britannica. But with the rise of AI, Nvidia saw an opportunity to pivot and use its advanced chips for computationally intensive applications such as edge computing and natural language processing. This has served as the basis for Nvidia's broader expansion into data center and cloud services. The company has recently taken to broadening its portfolio of related services — in May 2020, for instance, Nvidia closed a $6.9 billion acquisition of Mellanox, which develops software for managing enterprise data centers (famous for HPC networking Infiniband).

​

​

43

Public cloud providers are rapidly expanding their infrastructure in response to peak demand during the pandemic, and all the major cloud players use Nvidia's GPUs. The shift to remote work during the pandemic contributed to the largest-ever quarter-over-quarter (QoQ) spending increase for the cloud infrastructure sector — global enterprises collectively spent $34.6 billion on cloud infrastructure services in Q2 2020, marking a $3.5 billion increase from the previous quarter, according to Canalys. The three biggest providers — AWS, Microsoft Azure, and Google Cloud — are all investing in building out their cloud infrastructure. All three companies buy GPUs from Nvidia in cloud infrastructure, which they then lease to enterprise customers through public cloud services. In its earnings call for fiscal Q2 2021, Nvidia noted that its products such as the Ampere GPU, A100 are highly sought after by major cloud players since they lead the industry in several important AI performance benchmarks.

Business Insider Intelligence August 28 2020

Digital Science Center

�Big Data and Science

44

Digital Science Center

Digital Science Center

What is Cyberinfrastructure

• Cyberinfrastructure is (from NSF) infrastructure that supports distributed research and learning (e-Science, e-Research, e-Education)
• Links data, people, computers
• Exploits Internet technology (Web2.0 and Clouds) adding (via Grid technology) management, security, supercomputers etc.
• It has two aspects: parallel – low latency (microseconds) between nodes and distributed – highish latency (milliseconds) between nodes
• Parallel needed to get high performance on individual large simulations, data analysis etc.; must decompose problem
• Distributed aspect integrates already distinct components – especially natural for data (as in biology databases etc.)
• Cyberinfrastructure spans: High definition videoconferencing linking people across the globe; Digital Library of music, curriculum, scientific papers; Simulating a new battery design (exascale problem); Sharing data from world’s telescopes; Using cloud to analyze your personal genome; Analyzing Tweets…documents to discover which stocks will crash; how disease is spreading; linguistic inference; ranking of institutions; Enabling all to be equal partners in creating knowledge and converting it to wisdom

​

45

Digital Science Center

e-moreorlessanything

• ‘e-Science is about global collaboration in key areas of science, and the next generation of infrastructure that will enable it.’ from inventor of e-science term John Taylor Director General of Research Councils UK, Office of Science and Technology
• e-Science is about developing tools and technologies that allow scientists to do ‘faster, better or different’ research
• Similarly e-Business captures the emerging view of corporations as dynamic virtual organizations linking employees, customers and stakeholders across the world.
• This generalizes to e-moreorlessanything including e-DigitalLibrary, e-FineArts, e-HavingFun and e-Education
• A deluge of data of unprecedented and inevitable size must be managed and understood.
• People (virtual organizations), computers, data (including sensors and instruments) must be linked via hardware and software networks

46

Digital Science Center

47

The LHC produces over 50 petabytes of data per year of all varieties and with the exact value depending on duty factor of accelerator (which can be reduced simply to cut electricity cost but also due to malfunction of one or more of the many complex systems) and experiments. The raw data produced by experiments is processed on the LHC Computing Grid, which has some 200,000 Cores arranged in a three level structure. Tier-0 is CERN itself, Tier 1 are national facilities and Tier 2 are regional systems. For example one LHC experiment (CMS) has 7 Tier-1 and 50 Tier-2 facilities. CERN data center has 200 petabytes of data

This analysis raw data 🡪 reconstructed data 🡪 AOD and TAGS 🡪 Physics is performed on the multi-tier LHC Computing Grid. Note that every event can be analyzed independently so that many events can be processed in parallel with some concentration operations such as those to gather entries in a histogram. This implies that both Grid and Cloud solutions work with this type of data with currently Grids being the only implementation today.

Higgs Event

http://grids.ucs.indiana.edu/ptliupages/publications/Where%20does%20all%20the%20data%20come%20from%20v7.pdf

Note LHC lies in a tunnel 27 kilometres (17 mi) in circumference

ATLAS Expt. ATLAS is 45 metres long, 25 metres in diameter, and weighs about 7,000 tons. The experiment is a collaboration involving roughly 3,000 physicists at 175 institutions in 38 countries

Digital Science Center

48

9/7/2019

http://www.quantumdiaries.org/2012/09/07/why-particle-detectors-need-a-trigger/atlasmgg/

Digital Science Center

49

9/7/2019

http://www.interactions.org/cms/?pid=1032811

The inside of the RHIC (Relativistic Heavy Ion Collider) tunnel, a 2.4-mile high-tech particle racetrack at Brookhaven National Laboratory.

Digital Science Center

Tracking the Heavens

50

Hubble Telescope

Palomar Telescope

Sloan Telescope

Digital Science Center

Virtual Observatory Astronomy Initiative�Integrates Experiments over wavelengths

51

Radio

Far-Infrared

Visible

Visible + X-ray

Dust Map

Galaxy Density Map

Digital Science Center

52

Polar Grid

​

Lightweight Cyberinfrastructure to support mobile Data gathering expeditions plus classic central resources (as a cloud)

Digital Science Center

Genome Sequencing Costs

53

https://www.genome.gov/sequencingcostsdata/

Note performance of sequencing levelled off since 2015

Illumina NOVASeq 6000 can sequence up to 6 Tb and 20B reads in < 2 days. “Three years after its debut, NovaSeq makes deeper discoveries more accessible than ever. The 1000th sequencer was placed at the Whitehead Institute for Biomedical Research in Cambridge, Massachusetts in June 2020.”

​

Cost per Genome

Digital Science Center

54

The Long Tail of Science

80-20 rule: 20% users generate 80% data but not necessarily 80% knowledge

Collectively “long tail” science is generating a lot of data

Estimated at over 1PB per year and it is growing fast.

CSTI Meeting. October 2012 Dennis Gannon

Size of Dataset

Number of Research Groups

Digital Science Center

Data Intensive Activities

• Particle Physics LHC (bag of events of particles)
• Information Retrieval or web search (bag of words)
• e-commerce (bag of items with properties or users with rankings)
• Social Networking (bag of people with links & properties)
• Health Informatics (bag of health records, gene sequences)
• Sensors – web cams, self driving cars etc. (bag of pixels)
• Using
• Statistics (Histograms, Chisq)
• Deep Learning (Machine Learning)
• Image Analysis (including internet uploaded images)
• Recommender Engines (Bag of Ratings or properties)
• Patterns or Anomaly detection in graphs (linked data)
• On Clouds using MapReduce etc.

​

55

Bag=�Space

Digital Science Center

56

9/7/2019

http://www.wired.com/wired/issue/16-07 September 2008

Digital Science Center

The 4 paradigms of Scientific Research

1. Theory
2. Experiment or Observation
• E.g. Newton observed apples falling to design his theory of mechanics
3. Simulation of theory or model Supercomputers
4. Data-driven (Big Data) or The Fourth Paradigm: Data-Intensive Scientific Discovery (aka Data Science)
• http://research.microsoft.com/en-us/collaboration/fourthparadigm/ A free book
• More data; less models

57

Digital Science Center

More data usually beats better algorithms

• Here's how the competition works. Netflix has provided a large data set that tells you how nearly half a million people have rated about 18,000 movies. Based on these ratings, you are asked to predict the ratings of these users for movies in the set that they have not rated. The first team to beat the accuracy of Netflix's proprietary algorithm by a certain margin wins a prize of $1 million!
• Different student teams in my class adopted different approaches to the problem, using both published algorithms and novel ideas. Of these, the results from two of the teams illustrate a broader point. Team A came up with a very sophisticated algorithm using the Netflix data. Team B used a very simple algorithm, but they added in additional data beyond the Netflix set: information about movie genres from the Internet Movie Database(IMDB). Guess which team did better?
• Anand Rajaraman is a serial entrepreneur and �made big impact at Amazon. Datawocky is his blog
• http://anand.typepad.com/datawocky/2008/03/more-data-usual.html
• 20120117berkeley1.pdf Jeff Hammerbacher

58

Digital Science Center

�Big Data Systems

59

Digital Science Center

Digital Science Center

Background Remarks

• Use of public clouds increasing rapidly
• Clouds becoming diverse with subsystems containing GPU’s, FPGA’s, �high performance networks, storage, memory …
• Rich software stacks:
• HPC (High Performance Computing) for Parallel Computing less used than(?)
• Apache for Big Data Software Stack ABDS including center and edge computing (streaming)
• Surely Big Data requires High Performance Computing?
• Service-oriented Systems, Internet of Things and Edge Computing growing in importance
• A lot of confusion coming from different communities (database, distributed, parallel computing, machine learning, computational/data science) investigating similar ideas with little knowledge exchange and mixed up (unclear) requirements

60

Digital Science Center

Cloud Update Aug 3 2020

• Business Insider Intelligence
• Global enterprises collectively spent $34.6 billion on cloud infrastructure services in Q2 2020, marking a $3.5 billion increase from the previous quarter, according to Canalys. This represented the largest-ever quarter-over-quarter (QoQ) spending increase for the sector, though the overall growth rate continues tapering off given the overall scale of cloud spending.
• In 2022 95% of enterprise computing done on clouds

61

DIKW Process

• Data becomes
• Information becomes
• Knowledge becomes
• Wisdom or Decisions
• Community acceptance of results or approach important here
• Volume of bits&bytes decreases as we proceed down DIKW pipeline

62

Digital Science Center

63

9/7/2019

Digital Science Center

Example of Google Maps/Navigation

• Data comes from traditional maps (US Geological Survey), Satellites (overlays) and street cams
• Information is presented by basic Google Maps web page
• Knowledge is a particular optimized route
• Decisions (Wisdom) comes from deciding to drive a particular route

64

Digital Science Center

Classic Parallel Computing

• HPC: Typically SPMD (Single Program Multiple Data) “maps” typically processing particles or mesh points interspersed with multitude of low latency messages supported by specialized networks such as Infiniband and technologies like MPI
• Often run large capability jobs with 100K (going to 1.5M) cores on same job
• National DoE/NSF/NASA facilities run 100% utilization
• Fault fragile and cannot tolerate “outlier maps” taking longer than others
• Clouds: MapReduce has asynchronous maps typically processing data points with results saved to disk. Final reduce phase integrates results from different maps
• Fault tolerant and does not require map synchronization
• Map only useful special case
• HPC + Clouds: Iterative MapReduce caches results between �“MapReduce” steps and supports SPMD parallel computing �with large messages as seen in parallel kernels (linear algebra) �in clustering and other data mining

65

Digital Science Center

MapReduce “File/Data Repository” Parallelism

66

Map = (data parallel) computation reading and writing data

Reduce = Collective/ Consolidation phase e.g. forming multiple global sums as in histogram

Digital Science Center

Digital Science Center

�AI grows in importance

Transformation

67

Digital Science Center

Digital Science Center

Artificial Intelligence/Deep Learning

• Artificial intelligence (AI) forms a core part of end-user digital business strategies.
• Many end users today leverage the embedded machine learning (ML) capacities in systems such as SAP, SAS, Informatica and Spark to achieve their ML goals. For these workloads, traditional enterprise stacks can often deliver sufficient capabilities.
• However, deep learning (DL) leverages extreme compute power and large datasets to deliver unprecedented accuracy and efficiencies.
• Vendors are addressing the extreme compute requirements for DL training infrastructures by integrating a high density of compute accelerators (such as Nvidia GPUs) architected in scalable platforms.
• Other performance solutions that can scale out efficiently and �support a wide range of deep learning frameworks include �Tensorflow, Caffe2, Caffe, Theano, Torch, MXNet �and Microsoft Cognitive Toolkit (CNTK)

68

Digital Science Center

Transformation

• We will see in next 5-20 years more changes in Industry (and life) than in last 100-500 years
• Transformations can be looked in many ways
• Core Technologies related to
• New “Industries” over the last 25 years
• Traditional “Industries” Transformed
• Good to be master of Cloud Computing and Deep Learning
• Either Personally or through collaboration with others

​

69

Digital Science Center

AI First Engineering: Industries invented and remade through AI

• Core Technologies
• Digital transformation moving to AI Transformation
• Big Data
• Cloud Computing, software and data engineering
• Edge Computing and Internet of Things
• The Network and Telecommunications
• Apache Big Data Stack
• Logistics and company infrastructure
• Augmented and Virtual reality
• Deep Learning
• CNN for Images;
• Seq2Seq for Speech
• Everything else
• New “Industries” over the last 25 years
• The Internet
• Remote collaboration and Social Media
• Search
• Cybersecurity
• Smart homes and cities
• Robotics

70

Digital Science Center

AI First Engineering: Industries invented and remade through AI

• Traditional “Industries” Transformed
• Computing
• Transportation: ride hailing, drones, electric self-driving autos/trucks, road management, travel, construction Industry, Space
• Retail stores and e-commerce
• Manufacturing: smart machines, digital twins
• Agriculture and Food
• Hospitality and Living spaces: buying homes, hotels, “room hailing”
• Banking and Financial Technology: Insurance, mortgage, payments, stock market, bitcoin
• Health: from DL for pathology to personalized genomics to remote surgery
• Surveillance and Monitoring: -- Civilian Disaster response; Miltary Command and Control https://medium.com/mit-technology-review/military-artificial-intelligence-can-be-easily-and-dangerously-fooled-eb714cede48
• Energy: Solar wind oil
• Science; more data better analyzed; DL as the new applied mathematics
• Sports: including Sabermetrics
• Entertainment, Gaming including eSports
• News, advertising, information creation and dissemination, education, fake news and Politics
• Jobs

71

Digital Science Center

72

9/7/2019

http://www.kpcb.com/blog/2016-internet-trends-report

Digital Science Center

73

9/7/2019

http://sephlawless.com/black-friday-2014

No more malls?

Digital Science Center

No more malls?

74

2017 2018 Early 2019

Digital Science Center

Why is Technology so important in the AI/Internet Industrial Revolution?

• In the past, main industry use of computing was in areas like databases and not in many parts of leading edge computer science research
• Nowadays AI technology is central to large mass markets from Search to Media recommendations to Ride Hailing
• Now leading edge computing research is aligned with applications funded by billions of users
• This allows huge investment in core technology and also demands such an investment as the best service will deliver greatest reward to company.
• And these are often fights to the “death” with winner(s) takes all

75

Digital Science Center

Intelligent Systems ~ �Artificial Intelligence ~ Machine Learning

• 2017 Headlines
• The Race For AI: Google, Twitter, Intel, Apple In A Rush To Grab Artificial Intelligence Startups
• Google, Facebook, And Microsoft Are Remaking Themselves Around AI
• Google: The Full Stack AI Company
• Bezos Says Artificial Intelligence to Fuel Amazon's Success
• Microsoft CEO says artificial intelligence is the 'ultimate breakthrough'
• Tesla’s New AI Guru Could Help Its Cars Teach Themselves
• Netflix Is Using AI to Conquer the World... and Bandwidth Issues
• How Google Is Remaking Itself As A “Machine Learning First” Company
• If You Love Machine Learning, You Should Check Out General Electric�
• The trouble for all of these companies is that finding the talent needed to drive all this AI work can be difficult.

76

Digital Science Center

Some large areas: Google Trends last 5 years (Topics unless otherwise stated)

Search terms require exact match - topics are broader but sometimes are not available

77

SECURITY MAX IS 100

• Cloud Computing (Search Term)

• Big Data

• Computer Science (Field)

• Artificial Intelligence

• Security

Digital Science Center

Some growing areas: Google Trends last 5 years (Topics unless otherwise stated)

78

• Kubernetes �(Comp. App)

• Docker (Software)

• Amazon Web Services

• Artificial Intelligence

• Azure (Comp. App)

Digital Science Center

79

Digital Science Center

80

9/7/2019

Digital Science Center

81

Digital Science Center

Indeed.com Trends

82

Digital Science Center

�Conclusions and Jobs

83

Digital Science Center

Digital Science Center

Role of this class?

• Make you digitally savvy so you can take advantage of the AI/Cloud/Edge revolution whether as
• Domain Expert
• Data Scientist
• Data Engineer
• Software Engineer …….

84

Digital Science Center

Gartner on Data Engineering

•  Gartner says that job numbers in data science teams are
• 10% - Data Scientists
• 20% - Citizen Data Scientists ("decision makers")
• 30% - Data Engineers
• 20% - Business experts
• 15% - Software engineers
• 5% - Quant geeks
• ~0% - Unicorns � (very few exist!)

​

85

Digital Science Center

Conclusions

• The qualitative idea of Big Data has turned into a �quantitative realization as Cloud, Edge and Deep Learning
• Clouds are here to stay and one should plan on exploiting them
• Data Intensive studies in business and research continue to grow in importance
• Data Analytics: Everything is an optimization problem in a funny space solved using deep learning
• Growing employment opportunities in clouds and data related activities and so popular with students
• Opportunities in many of the most important companies from Facebook/Google/Microsoft/Amazon to General Electric and startups
• Students must be digitally/data savvy and flexible; technologies and opportunities are changing fast.

​

​

​

86

Digital Science Center

Big Data Ecosystem in One Sentence

• Use Clouds running Data Analytics Collaboratively processing Big Data to solve problems in X-Informatics educated in data science�
• X = Astronomy, Biology, Biomedicine, Business, Chemistry, Climate, Crisis, Earth Science, Energy, Environment, Finance, Health, Intelligence, Lifestyle, Marketing, Medicine, Pathology, Policy, Radar, Security, Sensor, Social, Sustainability, Wealth and Wellness with more fields (physics) defined implicitly
• Spans Industry and Science (research)
• This was the old mantra: �Now its deep learning on the cloud and at the edge

87

Digital Science Center