Paul
• Member for 7 years
• Last seen more than a month ago

## 76 Actions

comment

When I have a problem like this, I have found it helpful to pursue a hierarchical approach. I first use decision trees to segment out a large portion of the larger class. Typically there will be leaves in the tree with virtually none of the smaller class. The other leaves can then comprise the modeling dataset, which will be more balanced.

comment

Have you verified that the data values are unique in every file? Having them non-unique will cause a huge growth in the join size.

comment

The precise terminology is that the left graph is “wrong,” and the right graph is “correct.”

comment

This question is impossible to answer without some of the underlying information. Do you have a reference for how DL models fail vs. Asteroids?

comment

Why do you say that the given data indicates regression and not time series? These are not exclusive categories.

comment

Enough for what?

comment

To piggyback off of @DanCarter's comment, asking what ML algorithm to use is the wrong question. Think of the features you can engineer, and let that determine which methods to use (plural here is essential; you should never just try one method, unless the problem is extremely well understood). Some other possible features to try: distance from centroid (both absolute and relative to average point-centroid distance), distance from origin, angle the origin-to-point vector makes with an axis.

comment

100% accuracy always screams "target leakage."

comment

I am sorry, but you are not understanding the question at all.

comment

This is an explanation for what the different joins are, but the OP is looking for an example for why you would do one over the other, so this doesn't answer the question.

comment

How balanced is your datasets (ratio of 0's to 1's)? If it is very unbalanced, and there is not a sufficient relationship to the input, predictions staying below 0.5 can easily happen.

awarded
awarded
comment

Have you tried a periodic transformation of the inputs?

comment

Did you try any simpler models first?

comment

The clarification is asking about a recommendation engine, so it's not the same structure as what you give.

comment

That's not the kind of model this person is looking for.

comment

What exactly do you want your model to predict?

comment

Why an absolute difference? Keep the negative if it appears, although if you use the reference time as the point you are making a prediction, you never should have a negative difference, Since the negative difference would imply using future information to make a past prediction.