Business Analytics and Big data

Business Analytics and Big data

Welcome back this is the last lecture of this
section in fact from the beginning I told you 1st few lectures 1st few weeks 1st 4 weeks
in fact were Uhh devoted for knowing the functional e-business functional areas Next 4 weeks we
devoted including this class we devoted for knowing the technology Now over we did not
go to the details of this technology but knowing this overall technology is necessary for taking
business decisions And next 4 weeks we are going to talk about various decision support
systems Uhh which are resulted mostly because of these ICT technologies So right now we are going to focus on the
business analytics and big data Here we are going to learn what is business analytics
and role of big data In fact so far we have discussed about the technologies and various
subsystems Now data is generated from every source from the transactions when you carry
out transactions starting from a transaction level you collect the data That data we know
is used for management decision support your management information system and decision
support system uses data to take Uhh tactical level decisions Then this data is further refined and used
at the strategic level as well And in last few lectures we saw not only these transition
data we also we are also getting location data Not only location data you are also getting
the data about the quality of the product So this huge amount of data that we are getting
about the transactions about the products about the locations etc what are we going
with going to do with this data Because now we have become informed with the available
of the data while discussing about the difference between data and information we know that
data is something which just exists Unless and otherwise we have appropriate tools
and techniques and models to use this data this is basically useless So here is the role
of analytics What is analytics analytics is the use of data information technology statistical
analysis quantitative methods and mathematical or computer-based models to build models to
help managers gain improved insights about their business operations and make better
fact-based decisions So analytics is about making fact-based decisions using the data
the technology and responding mathematical models So broadly we can classify these analytics
into 3 categories descriptive which is about data visualisation and statistical descriptions
predictive analysis which like you are forecasting predicting descriptive analysis Uhh prescriptive
analysis like telling you Uhh what is the right way of doing something is through optimisation. This in fact in one of the earlier classes
we have already told little bit what are various types of questions that are answered by these
analytics Now coming to this disruptive analysis the question such as what happened in the
past what is right now happening what actions are needed what exactly is the problem what
actions are needed and getting answers to this are your descriptive analytics So here
the tools are use of ad hoc reports dashboards data warehouses genetic alerts are some of
the enablers What is the outcome outcome is a well-defined
business problems and opportunities Now analytics can be predictive which answers questions
such as why something is happening what will happen next and why is it happening why will
it happen The answers to these questions can be found by using data mining text mining
web data mining and forecasting tools Again accurate projections of the future states
and conditions which Uhh helps in forcing certain business situations and take some
decisions are the outcome of this kind of analytics The prescriptive analysis as the name indicates
prescribes something so the question such as what should I do why should I do it what
is the best that can be that can happen what if if we try this and so on are answered here
So after this after knowing about these 3 descriptive Uhh 3 types of analytics that
this descriptive predictive and prescriptive let us try to see one example Let us say this example is about some retail
markdown decisions in fact most department stores clear their seasonal inventory by reducing
prices Now the question is when to reduce the price and by how much Through descriptive
analysis we can examine we can view descriptive analysis is basically viewing the data To
examine the historical data for similar products prices units sold advertising that has happened
we can visually see it and think of taking some decisions Then next step is we can use
some kind of predictive analysis for predicting sales based on price and taking some decisions
accordingly Next we can use some kind of prescriptive
analysis to decide what is the best pricing and what is the advertising strategy that
can maximise my sales revenue So for the same situation now we can use descriptive analytics
predictive analytics and prescriptive analytics. Okay now we coming to our supply chain situations
if we would like to have such kind of analytics in the supply chain we need some data management
resources for acquisition and management of the data from your RFID ERP and other databases
We can use some kind of analysis which includes data mining other analysis of course are there
We need to we need to have IT based supply chain planning resources support of this and
support of And we can also help in performance management for statistical process control
for managing your 6 Sigma process and all Now as we have been talking about this analytics
is about making informed decisions informed fact-based decisions Now when it comes to
fact-based decisions the facts are basically what about the data And as I have told the
data is no more limited to transaction data volume of the transaction data is already
increasing but the kind of source that we have discussed more recently like that of
sensors IOT your RFID and other IOT devices etc. GPS and so on are generating huge amount of
data Those data are different from your transaction data However there is no standard definition so
far for big data but this definition I got from some other source In fact some of Uhh
the facts I have got this from the source itself This big data is a Uhh is data whose
scale diversity complexity requires new architecture techniques algorithms and analytics to manage
it and extract the value and hidden knowledge from it. There are as they say there are some 5 Vs
associated with big data in fact some of them countdown on 3 Vs some of them are 4 Vs and
5th V is basically knowing about whether you have value associated with the data or not
But anyway in technical terms people usually use 4 Vs and the way the Vs are written are
also different because the literature is not yet standardised but anyway the 1st important
characteristics of this big data is this volume In fact there is almost 44 times Uhh the data
increased between a very small period that is from 0.8 Zettabytes to 35 Zettabytes it
is a huge change This volume this data volume is as this Uhh
data generating sources keep generating the data this volume is increasing exponentially
The 2nd characteristic of this big data is variety It is comes in various formats types
and structures For example your RFID data you GPS data have other sensor data data from
the customers Uhh your which comes from the social media etc they are in different formats
someone in text numerics image audio video sequences timeseries and so on Some of this data are static and some of them
will be streaming they will be coming continuously their velocity is very high Then though they
are coming from various sources finally data fusion has to happen to take certain decisions
So a single application in the scenario big data scenario can be generating or collecting
data from many many different sources Then the 3rd characteristics is actually velocity
which is about how frequently the data gets generated For example think about your RFID
data GPS data they come depending on your device’s capability they come in fact continuously
they come in even you can adjust it to milliseconds or seconds or whatever based on your choice
Your stock price data it comes so frequently varies seconds to seconds variation you can
see So if you do not have proper online data analytics tools as soon as this data comes
you should be able to take decisions if you do that in fact you lose many business opportunities Examples include your E Promotions: based
on your current locations your purchase history what you like send Promotions right now to
store next to you Healthcare monitoring: sensors monitoring your activities and body Any abnormal
measurement requires immediate reaction These are the 3 Vs volume velocity and variety which
are associated with how complex data how speedily it is coming and what is the volume If you
look at this diagram while from ERP you were getting the data in megabytes from CRM you
are getting the data in gigabytes from where you are getting the data in terabytes and
from your RFID sensors you are getting the data in petabytes You can see other details from here what are
various applications of this data what are various sources of this data and as we increase
along this line there is increase in decision-making often very complex decision-making requires
increased in not only in data volume in data variety and complexity Then some even Uhh and a 4th V that is called
veracity which is about about uncertainty and inconsistency associated with the data
and approximately solving them So you have these 4 dimensions data at rest which basically
is terabytes to exabytes of existing data to process Data in motion this is the streaming
data this has milliseconds to seconds to respond You have many forms in structure unstructured
text and multimedia data You have data in doubt which is uncertainty due to the data
inconsistency incompleteness ambiguity latency deception model and approximation and so on Now in order to harness the big data you have
many technologies Online transaction processing online analytical processing 1st one is related
to database in fact you you remember the lectures that we did while managing out data resources
in fact many of these things we have covered a little bit more detail And about real-time
analytics which has basically evolved because of this Uhh high velocity big data Now who all are generating social media networks
scientific instruments mobile devices sensor technologies and network Now what are the
driving what is driving this big data is ad hoc querying and reporting which is for business
intelligence data mining use of data mining technique which uses structured data typical
sources small to medium-sized data sets then relative analysis in data mining which requires
optimisation predictive analytics complex statistical analysis data from various sources
large datasets and more of real-time data What are technologies From very deep fast
data to large volume of the data you have various technologies you can just name this
because Uhh knowing about these technologies simply does not I mean naming this technology
does not make sense But it is basically what this diagram indicates is
how Uhh what are the what are the technologies for various types of data and what kind of
insights you get and what kind of analysis you do In fact some of these analysis and
a few more data sources and little bit more on this analytics part we will be knowing
while discussing about the decision supporting Uhh in e-business specifically we will be
talking about decision-making situations which are resulted because of because of new phenomena
that is happening in the business world Now in the supply chain literature big data
is coming from many sources while tracking the products from GPS and RFID in manufacturing
in manufacturing so far in we have not have any decision on this The data the parts themselves
are talking about parts are now talking to each other and they are sending data to some
centralised server these are called cyber physical systems In fact Industry 4.0 is all
about it this is about machine to machine communication as a result huge amount of data
is getting generated Next is prevailing sales and marketing while
analysing weblogs understanding customer reviews again you encounter big data This is a more
Uhh detailed example of various during various datatypes and based on this volume velocity
and variety For example sales you have more detailed price quantity item time of the day
a date customer where the sales is actually happening This is velocities from monthly
weekly to daily and sometimes it is hourly Variety direct sales data distribution recruited
field data Internet sales data International sales competitors sales data is coming from
various sources this is variety In case of consumer you have more details
about the items browsed bought frequency dollar value and timing Then velocity is coming from
the click through to the card users credit card users Then you have variety which includes
software identification emotion detection like his likes tweets etc. his correct reviews
and so on Inventory perpetual inventory by style color size this is can be velocity can
be found monthly update to hourly monthly update to hourly update data is coming from
warehouses food Internet store when the inventories and so on Location and time data sensor data to detect
locations better inventory control and all Velocity is frequent updates within the store
in transit and so on It is not only where but what is chosen who moved it path future
path and so on Then this big data provides various opportunities
to improve the supply chain operations for demand forecasting by linking the real-time
sensors to machine learning Uhh sensors to machine learning algorithms barcode checkouts
and in case of Walmart RFID chips already exists Then this enables real-time responses
about the demand Warehouse design and location you can design system for optimality it is
a classical operations research trouble of course but it can use network analysis to
be more complete Supplier evaluation selection it can consider more factors and more up-to-date
data Selection of transportation nodes real-time
truck and rail assignment so such systems already exists and as more data comes in but
algorithms are warranted Companies are using this big data for their advantage Both it
includes both start-ups and established companies For example UPS it tracks packages avoidance
vehicles and tours them Then Schneider International for Uhh trucking for trucking it gets the
sensor locations driver behaviour etc So these are some of the uses the companies are already
making In grocery stores in stores like that of big
stores like that of Walmart by airline companies by trucking companies are all happening or
getting revolutionised because of this big data So with this we finished this lecture
from next class onwards we are going to Uhh learn about the decision support modules in
e-business thank you very much

You May Also Like

About the Author: Oren Garnes

Leave a Reply

Your email address will not be published. Required fields are marked *