Introduction To Machine learning
Without any great hope in its results NOTHING HAPPENS!
Imagine one day you wake up in the morning and something insane happens, you don’t really know what it is but you’re lucky because you recognize it before the game is over.
This is not something easy to see cause it’s hidden between you and your life.
The world is changing faster than ever due to science and technology.
The way business operates is transforming and now internet open up a big window for those business owners.
There is a secret here behind
- winning the election
- diagnosing illness very precisely
- estimate how much things costs
- predict the football winners
- figuring out the new book,new music, new movie, new clothes.. similar to what you are looking for
- know whether customer adores and satisfy with your product/service due to rate reviews
- let your car observe what you do as a driver to learn how to drive itself
- discover sophisticated scientific research and new approaches of dealing with theories and theorems
- knowing whether your card number was stolen or not
- make fast trading decisions if you are in finance market
- discover new correlations between different branches, classes, subjects and groups to know when you should act, what you’re going to do next and which behaviour you have to adapt in a particular situation
- feel understood anytime you check your phone, it knows you very well
Nothing of All these situations happen until Machine Learning Algorithms come here.
What is Machine Learning
So, Machine learning is a branch of science that means allow machines to learn from data without anyone tell them what to do explicitly.
As you know Computers aren’t creative, they just execute.
But what if you tell them to do is to be creative then you get Machine Learning.
Almost we are focusing on results, how much customers up this year, how much money do I double 6 months from now, how much gain can I have after closing this deal, how could I get results in my marketing campaign???
And Machine Learning helps us predicting the results of our actions to achieve certain goals.
But How?
It is all about Data.
The more data you have the better you get.
there is input and output and between there is a process.
Modeling the process is being able to get reasonable and more accurate output through the data you put in.
How to choose the right Algorithm?
To get the outcomes you need, you have to choose the right algorithm meaning which algorithm corresponds to the situation, for that you have to:
Analyze the situation :
by understand your data and categorize it, is it classification algorithm meaning the output is whether 0 or 1, for example spam detection (spam or not spam), is it regression algorithm that predicts the output values based on input features like estimating the final price of a house based on features of multiple houses.
Process the data:
it means preprocessing data in terms of :
- integrate data_ same format
- clean data and dealing with missing values
- normalized data and scale it for example age is between 0 to 100 years but the number of population in countries is in range of Millions then you would normalized this range to get it match with age feature.
- dimensionality Reduction: reduce the dimension of features because most of them are correlated , instead of 3D that is hard to visualize you simply reduce it to 2D.
Complexity:
keep it simple, less complex and eliminate multiple Algorithms under one interface.
Example of 3 basics algorithms:
There are multiple types of algorithms, let’s look at 3 basic algorithms:
- Naive Bayes:
Based on Basyes’s theorem used to predict the class of unknown data sets. It can be used to classify a news article and check at the same time positive or negative emotions of a user.
- Logistic Regression:
Also is a classification learning algorithm (binary result), it’s a linear method using the logistic function, logistic regression can be used for cancer detection problems.
- Decision Tree:
As there name indicates, it’s a tree representing to solve a problem in which each leaf node corresponds to a class label that contains conditional statements, used often in decision analysis to help identify a strategy most likely to reach a goal.
Artificial intelligence vs Machine learning :
Machine learning is sometimes confused with Artificial intelligence.
Machine learning is a subfield of AI, the goal of AI is to stimulate natural intelligence to solve complex problem and increase chance of success. And ML focuses on learning from data on certain task to increase accuracy but not necessarily care about success.
Statistics vs Machine learning :
Machine learning requires statistical thinking.
Machine learning adopts many of statistics methods, but it will never replace Statistics, why?
Both Statistics and Machine learning create models from data, but for different purposes.
In Machine learning, the predominant task is predictive modeling that will predict new examples.
But Statistics designed to characterize the relationship between data and our outcome target not to make predictions about future data.
Why businesses embrace Machine Learning
Why company X is much more than company Y? They both make their money from the same sources and both use machine learning in there operations, but the only difference here is learning Algorithms of company X are much better than company Y .
Machine learning is a big part of growing companies faster. They look at it as a gold asset.
It has a huge part of knowing more their customers rather than just focusing on things.
knowing their needs, who they are, what they’re looking for, their moods, their feelings, their feedbacks and so on..
Netflix may have one hundred thousand DVD titles in stock, but if customers don’t know how to find the ones they like, they will default to choosing wrong one. It’s only when Netflix has a learning algorithm to figure out your tastes and recommend the best fit to you.
Once the companies succeed in this level of making customers feel understood, the number of sales increase and companies thrive efficiently.
In 2019, Google net worth is estimated to be $336 billion, and this is mainly due to the advanced learning Algorithms the company has.
Whoever has the best algorithm and the most data win. Whoever has the most customers databases wins new more customers.
It’s not just understanding customer better, companies can apply machine learning to every aspect of their operations.
And companies without machine learning can’t keep up with one that uses it!
“A baby learns to crawl, walk and then run. We are in the crawling stage when it comes to applying machine learning.” ~Dave Waters
Machine Learning is a huge opportunity for either consumer or producer, it revolutionize the world of business and improve other’s life. But we should recognize that we are just in the first stage of machine learning, many more things will come…
