1
$\begingroup$

I am currently new in Data Science field especially in ML domain. There are various approaches to a problem ie. Supervised, Unsupervised learning. I want to know what is the approach and how to determine using exactly which algorithms in real life test cases.

Even if a particular algorithm is selected, how do we know other algorithm might have better accuracy and precision than this one

$\endgroup$

closed as too broad by Icyblade, Dawny33, ncasas, Sean Owen May 30 '17 at 14:49

Please edit the question to limit it to a specific problem with enough detail to identify an adequate answer. Avoid asking multiple distinct questions at once. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

5
$\begingroup$

Try as many models as you have time for and pick the best one, otherwise use your intuition to narrow the candidates. If you have no intuition, start with a simple model and move to more complex models as performance requirements dictate. Often there are reasons beyond pure performance metrics to choose a model in the real world, such as simplicity, interpretability, and speed.

You will not, however, have to pick between supervised and unsupervised learning, for these are drastically different things. If I ask you to group all similar items, how are you going to apply supervised learning?

You might like this map about choosing the right estimator.

$\endgroup$
  • $\begingroup$ Thanks @Emre, the map is really helpful. I was looking for something similar. $\endgroup$ – parth May 26 '17 at 7:10
1
$\begingroup$

Basically it all depends on , which type of problem you are trying to solve. "Even the most experienced data scientists can’t tell which algorithm will perform best before trying them." So Here some parameter on which you can decide which algorithm you pick up for your machine learning problem.

1. Accuracy

2. Training time

3. Linearity

4. Number of parameters

5. Number of features

Microsoft released a PDF cheatsheet of what machine learning algorithms to use, when. Download the cheatsheet (PDF) from its companion blog post titled “Machine learning algorithm cheat sheet for Microsoft Azure Machine Learning Studio“

$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.