Machine
Learning or Artificial intelligence :
- Machine learning is the leading edge of artificial
intelligence.
- It is the subset of artificial intelligence where
machines use algorithms to interpret data from the world to predict
the outcomes. To learn from sucess and failures.
- As machines infiltrate accounting tasks to take over
more mundane and repititve tasks, it will free up more accountants
and book keepers to spend more time using their professional
knowledge and interpret data which will be provided to their
clients.
Positive impacts of
implementation of artificial intelligence or machine learning to
counteract issues :
a) Casual agency :
- When artifacts are said to be in the agents sense the
emphasis on the agents efficacy. This type of agency fills the gap
since too little attention has been given to the ways in which the
technology shapes what happens in the world.
- Treating artifacts as the agents properly frames them
as th significant constituents of the human world.
b) Intentional agency :
- Intentional agency involves the intentional actions
that are used by the intentional agents which are the entities that
act intentionally.
- Casual agency and intentional agency share the element
of casual efficacy.But in intentional agency agent's intention
begin the chain of casuality.
- Intentional actions of top management come into play
for resloving agency issues.
c) Underlying data are often the source of
bias:
- Underlying data rather than algorithm itself are the
most important and main source of issue.
- Models are trained on the data containing human
decisions on the data that reflect second order effects of
societial or historical inequities.
- For example, word embeddings ( a set of natural
language processing techniques ) may exhibit the gender stereotypes
found in the society.
d) Budget constraints in the training phase
countered :
- Classification problems for which feature labels are
unknown in the trianing phase which are commonly countered in
crowdsourcing.
- The algorithm can acquire the label of a desired
feature and each label has a fixed cost.
e) Budget constraints at testing phase
countered:
- Budget constraints at testing phase are countered using
feature labels for the test.
- crowdsourcing prediction markets,medical
diagnosis,sensor data aggregation are the new prdictions used in
the testing phase to acquire the data required for the specific
instance and are used to reslove the issues.
f) Risk Assessment :
- Machine learning could facilitate risk assessment
mapping and pulling the data from every project of the company had
ever completed to compare it with propsed project.
- Assessment of risk helps in counteracting the issues of
security through fund accounting.
g) Artificial intelligence for financial security
:
- Artificial intelligence provides an adaptive technology
that can analyze the activity of attackers.
- Artificial intelligence with the potential to provide
customized security systems have different security
requirements.
- Artificial intelligence is fully capable of unertaking
data analysisin an automated fashion to identify and flag
anomalies.
- Through amchine learning algorithms artificial
intelligence can predict potential attacks on the current
data.