Please enable JavaScript.
Coggle requires JavaScript to display documents.
Machine Learning (weaved into everyday-life with unseen threads (You…
Machine Learning
currently at the core of many if not most organizations’ strategies
these tools are integral for knowing the customer, streamlining existing operations, and managing risk and compliance
Being able to evaluate what happens when a model produces high-stakes decisions is a critical reason for learning this content.
is the strength of these relationships, and it becomes a “definition,” if you will, of what drives
customer retention
is a set of relationships between the features and the target
can provide a probability (from zero to one), in a matter of milliseconds, of whether or not this customer is likely to be retained
there will be two classes of organizations: those who use machine learning to transform their capabilities
and those that do not
The organizations that use machine learning effectively and survive will most likely focus on hiring primarily individuals who can help them in their journey of continuing to derive value from use of machine learning
understanding how algorithms can and are used in business will become an essential core competency in the 21st century
will affect people and organizations no less than the industrial revolution’s effect on weavers and many other skilled laborers
machine learning will automate hundreds of millions of jobs that were considered too complex for machines to ever take over even a decade ago
driving, flying, painting, programming, and customer service, as well as many of the jobs previously reserved for humans in the fields of Finance, Marketing, Operations, Accounting, and Human Resources.
weaved into everyday-life with unseen threads
You yourself are the subject of machine learning hundreds of times every day
Your social media posts are analyzed to predict whether you are a psychopath or suffer from other psychiatric challenges
your financial transactions are examined by the credit-card companies to detect fraud and money laundering
each of us log into a unique, personalized Amazon Marketplace tailored to our lives by their highly-customizing machine learning algorithms
our brains contain an untold number of models that help us predict the outside world
If our forebears had not developed such abilities, we would have long since died out.
even judges contemplating the level of punishment for convicted criminals make their decisions in consultation with machine learning models
the majority of stock trades are now done through machine learning
work wonderfully for most large-scale problems
if you know where to look, you’ll find small mistakes
Finding these slips and learning from them could become one of your defining strengths as an analyst
“a field of study that gives computers the ability to learn without being explicitly programmed”
already drive hugely disruptive events for humanity, but the best is still to come
computers to learn to solve a problem by generalizing from examples (historical data), avoiding the need to explicitly program the solution
split into two types:
supervised
a case of the data scientist selecting what they want the machine to learn
a model is trained to split people into two groups: one group, “likely buyers,” and another group, “likely non-buyers.”
historical data to understand relationships in the world that can then be used in predicting future outcomes is the key central concept that is transforming the world
relative clarity of the idea and the ever-expanding business potential available in its implementation
is where most of the benefit of machine learning lies, and will be our focus this semester
unsupervised
leaves it to the machine to decide what it wants to learn, with no guarantee that what it learns will be useful to the analyst
constantly searching for relationships between features and a target, often as naively as a toddler, but with enough data that the model will outperform most human adults.
The target (often an outcome) is what we are trying to understand and predict in future cases
is
in the coming decade
's current revolution
including
A popular meme once decried that you should “always be yourself, unless you can be Batman, then always be Batman.” Akin to this: if you can find a way to be a data scientist, always be a data scientist (especially one as good at his or her job as Batman is at his), but if you cannot be a data scientist, be the best self you can be by making sure you understand the machine learning process.
for example
but
algorithms
was defined by Arthur Samuel as
enables
is
can be grossly oversimplified as:
for example:
due to the
uses
are
A model: