.What is the
major limitations of algorithms?
There are two big
limitations of algorithms which we discussed above.1.They are literal and 2.
They are dark boxes. In an organizational world, there is high level use of
algorithms but at the same time people should have proper knowledge of all
things all over? Algorithms are very literal. They will only follow that we
said to them and we said to do them even it’s wrong or right.
use commonsense even we said to do true or false. For example- David
Lazar, a computer scientist at Northeastern University, is In a recent Science article he said:
that human lives are regulated by code is hardly a new phenomenon.
Organizations run on their own algorithms, called standard operating
procedures. And anyone who has been told that “it’s a rule” knows that
social rules can be as automatic and thoughtless as any algorithm.algorithms
can give us result of our questions but it would be right or wrong according to
situation it’s hard to say.
They are dark boxes. They will provide us data and details
of our every input but why they reached at this result they will not tell thing
because they do not have sense they are artificial Intelligence which will not give reason for their result. For
example-We search online that would
be inflation rate of economy in Australia in 2020 and we get 22.3%.we
got our answer but not reason that how and why it will go to 22%.it is just a
prediction of algorithms on the basis of old data.
Q3.Why algorithms are predictor not advisor?
algorithms create improved predictions than people when they cannot do. We can
get result with algorithms but they cannot advise to do anything. It’s totally
depending on our decision. This is a prediction, not advice.it works only as a forecast
since there are many other issues that relate with small tweets that make them real.
This is the most important reason why it flops as advice. They utilize existing
data to create expectation about could be happen with a slightly diverse
setting, population, time or question
but at the same time algorithms cannot suggest to take any action. For example-we
want to buy shares of commonwealth bank and firstly want to predict future of share
market and we get positive result but it does not mean that algorithms
contribute any advice that we should invest or not. Algorithms can give us
expectation but not advise to participate or not, it rest on our suitable
Q4. How important is to choose right data inputs in
algorithms before to get any result and which things we should keep in
make system smarter, it is must to put right instructions to system and how
much data is possible to provide to get any result.it is very helpful for best
and accurate result. To choose right data resources, some are very
indispensable stuffs which we must retain in mind. That is mixture (range) of data and also broader data.
RANGE OF DATA: To get reliable and accurate result
diversity of data is most important aspect of algorithms because they give us
result on the basis of our provided data so how much long data we will provide
to algorithms it will help them to make our decision more better and correct.in
the sense, extra data will make extra prediction power but diversity of data,
data should not match with each other, it should be additional one.
BROADER: Every data or word is sign. Every
additional detail and data which we learn and give to algorithm is supportive
for result and would be like one more sign, while growing the length of data
will expand prediction of procedures.But with broader range of figures, data
management is also very important. Data management improves management system
but better timing of data is very important. These are the some very essential
things that should be on first place when thinking about to put right data for
study we learnt that before to use use any artificial power or technology
measurement of its limitations is really important part. To know its
consequences can reduce some future problems very easily. Because every
technology has some merits and demerits and we cannot ignore its demerits in
front of its benefits. Use of algorithms is very beneficial and we cannot
reduce it but we can overcome t its problem with the better use of it.
Chief Executive Douglas Merrill told Gage, “Data scientists need to verify
whether their findings make sense. Machine learning isn’t replacing people.”
Part of the problem is that most machine learning systems don’t combine
reasoning with calculations. They simply spit out correlations whether they
make sense or not
CONCLUSION:-From the overall
analysis from article we came to know that artificial power is more helpful and
fast rather than human power but when we will neglect its limitations it can
create delay and error in wok so its essential that we have proper knowledge of
its consequences before to use it.with the help of algorithms it’s very
important to follow your brain as well because they can help us but cannot
suggest anything. Absence of proper knowledge to use it can affect lots of
things in one stream. They can show us
path but it is wrong or right it’s depend on our knowledge and understanding.