broccoli
• Member for 9 years, 5 months
• Last seen this week

A write up of probability tutorials and related puzzles along with R code for learning. Hope it helps

This is tricky. Often times, the actual position of a word is not that useful in enhancing any kind of recall you want to do on the document. But you could, for instance, create features around them ...

If you are looking to bulk up on machine learning/data mining I would strongly urge optimization/linear algebra/statistics and probability. Here is a list of books for probability. Hope that helps.

This a tough one without revealing more about the data. I would say your best bet is to create features out of the plain text file based on subjective data. For example some features could be 1) How ...

For an Naive Bayes classifier, categorical values are the easiest to deal with. All you are really after is P(Feature | Class). This should be easy for the days of the week. Compute P(Monday | Class=...

A simple explanation for this would be that there are 3 ways in which sleeping beauty can wake up two of which are from a Tails toss. So the probability has to be 1/3 for a heads every time she wakes ...

Have you tried plotting this using ggplot2 in R? It has a nice semi transparency feature with the Cairo package which makes guesstimating the mean for such residual plots easy. For example you could ...

The solution might be easier than you think. Why not try something like the following (denote as $x$ the ratio of number of workers to square footage) $log(sales) = \alpha_0 + \alpha_1x + \alpha_2x^{... View answer 1 votes A lot would depend on the data set itself. From your graph it appears that the RF(or NN) is missing on training some information. In such cases, you must almost definitely do what you are proposing. ... View answer 1 votes It is possible to handle such situations. For example you could try and use a simple exponential moving average to predict the next sample point, and then use it again to predict the next to next ... View answer 1 votes Here is a write that describes something very similar to that. The Bayes approach is the right way to proceed. View answer 1 votes A 75% win probability is definitely a good edge to have. To answer your question, yes it is possible to generalize for n: 100 - n. The Kelly criteria is a good way to go, but as is pointed out in this ... View answer 1 votes You can use numerical values quite easily. In the term P(Feature|scam=Yes) you could put a gaussian distribution or any other empirical distribution from training data (for e.g. sort the data, create ... View answer 1 votes What you ask is not exactly a question. I can point you to this blog which has a few questions worked out. HTH View answer 1 votes If a classifier predicts a certain class with a probability, that number can be used as a proxy for the degree of confidence in that classification. Not to be confused with confidence intervals. For ... View answer 0 votes If the original models did not consider conditional probabilities you are only left with the option of treating both these probabilities as independent.$ P(not cancel) = 1 - P(cancel) $. Here$P(...

Reading up books on neuron models is a good start, but it may not be the best way to start your journey to be an expert in probability and statistics. Some of the concepts in that area may leave you ...

You can try casting the problem as a regression problem, wherein you are trying to predict the probability score. Or perhaps something simpler like replicating more training instances for cases where ...

Here is a good list to learn the art of probability & statistics. Here is another set to learn monte carlo methods. Note, you are better off getting a good grounding in statistics and probability ...

Odds ratio for a success rate of $p$ is defined as $\frac{p}{1-p}$. A nice feature of this ratio is the plain English explanation you can do with it in a betting scenario. As an example, assume $p = 0.... View answer 0 votes I liked the book "The practise of Business Statistics" as a good verbose introduction to the application of creating models with some real world data with real world problems. The mathematics in the ... View answer 0 votes Here is a good list of books on time series analysis. Note that there is a lot of difference amongst books that cater to people of different backgrounds (economists/engineers/statisticians). hth View answer 0 votes Here is a good collection of books in probability and statistics including R programming. Stay away from "dummies" or "idiots" guide type books because (I think) they needlessly dumb down concepts in ... View answer 0 votes Learn R. Absolutely, definitely do so. R is open source and I think will soon become the de facto standard for the statistics & machine learning. You also need to learn a scripting language and I ... View answer 0 votes Here is a list of books. Puzzles/riddles are a great way to instil an interest in what mathematics/statistics can do. Real life examples help too. View answer 0 votes The t-test is apt here. There is no dependency on the range involving zero. So long as$t = \frac{\delta \mu}{s}\$ is within range (decided by the chosen p-value) for the given number of degrees of ...

Very nice question. A thing to realize upfront is that machine learning is both an art and science and involves meticulously cleaning out data, visualizing it and eventually build models that suite ...

Here is a good write up on generalized linear regression. The code is done in R and it explains how they work. CRAN also has a package glmnet which does this for you but can be a bit unwieldy to use ...

You are mostly on the right track here. Standardization is generally a good practice because lots of simple mainstream approaches you would use (like linear regression) assumes you have a normal ...

Here is a good collection to buy. I like the "Bundle of algorithms in Java", it gives straight out implementations/examples as does "Machine learning, practical tools and techniques" which is also a ...