Stack Exchange Network

Stack Exchange network consists of 174 Q&A communities including Stack Overflow, the largest, most trusted online community for developers to learn, share their knowledge, and build their careers.

Visit Stack Exchange

Questions tagged [predictive-models]

Predictive models are statistical models whose primary purpose is to predict other observations of a system optimally, as opposed to models whose purpose is to test a particular hypothesis or explain a phenomenon mechanistically. As such, predictive models place less emphasis on interpretability and more emphasis on performance.

0
votes
0answers
10 views

Bias and Variance in underspecified models

Galit Shmueli (2012) introduces in her paper "To Explain or to Predict" the biases and variances of correctly and underspecified predictive models. The correct model is $f(x)=\beta_1x_1+\beta_2x_2+\...
0
votes
0answers
18 views

Anti-Correlate Data [on hold]

In the attached image, I want to identify the start and end time of certain periods where those selected variables are anticorrelated using some algorithm (machine learning?), which one do I choose? I ...
1
vote
1answer
37 views

Predicting whether house is sold: regression or classification

I am new to machine learning (I am currently following the Udemy course machine learning from A-Z). Basically, I would like to reproduce the following analysis (https://www.datasciencecentral.com/...
0
votes
0answers
12 views

caret predict.gbm function error: factor has new levels [on hold]

Similar to THIS question, I am getting an error in my prediction stating that one of my factors has new levels. I am using the gbm method from ...
0
votes
0answers
8 views

Can overfit happen in spite of validation and what to do with it?

Let's consider a standard situation where we need to find a predictive model. We train all the available model using a training data set. We validate all the trained model using a validation data ...
0
votes
1answer
12 views

How does train-validation-test procedure deals with the sampling error of the accuracy measure?

Let's consider a standard model selection procedure: We have N different untrained models (for example linear regression, neural network, decision tree and so on). We use a data set A to train each ...
0
votes
0answers
28 views

High dimensional regression overfitting

Consider the linear regression model \begin{equation} \boldsymbol{y} = \boldsymbol{X}\boldsymbol{\beta} + \boldsymbol{\epsilon} \end{equation} where we assume $\boldsymbol{X}$ is $n$-by-$p$, with $p &...
0
votes
0answers
27 views

ML - Text input data

I'm trying to explore an use-case in ML but stuck at a point. May i please request your advise please. Have a service desk web application for logging tickets, which is essentially a form having ...
4
votes
1answer
64 views

How much of a problem are autocorrelated residuals of a binary GAM (Generalized Additive model)?

I'm trying to predict high or low crime rate in regions (binary 1/0 response variable) using a range of socioeconomic variables. Im doing this with a panel dataset with 300 regions over 17 years (2006-...
1
vote
1answer
31 views

How can I create a meaningful weighting for RMSE?

Background I should start off by saying I am not a mathematician and please excuse simple/stupid mistakes! The goal of my exercise is to find the “best-fitting” model for the purpose of prediction. ...
0
votes
1answer
21 views

What is the formal name of the following problem: predicting next output based on previous values, NOT a sequence

I have a system producing an output periodically, I would like to build a model to predict the next entry. There is no sequence relation between the output values, order doesn't matter. The only ...
0
votes
1answer
13 views

How to do Predictive Modeling for Different Companies

There is Company X Company X wants to predict if its opportunities will Win or Lose I trained my model based on the data of company X and did the prediction of new opportunities Now Company Y and ...
1
vote
1answer
25 views

How to improve a model that is consistently underestimating

I've been trying to predict house prices (real data from my country) and I noticed that initially, errors are centered around zero, but around the $2,500,000 mark, the model starts underestimating ...
0
votes
0answers
18 views

With omitted variables is OLS estimator still the best linear predictor?

Suppose the true model is $$y = \beta_0 + \beta_1 x_1 + \beta_2 x_2 + \epsilon$$ where $x_1$ and $x_2$ are correlated and $\epsilon$ is white noise. I omit variable $x_2$ and apply OLS to estimate $...
0
votes
0answers
8 views

How to use the multilayer perceptron to predict the price of an action with and without news?

I have a multilayer perceptron that predicts the value of a stock market given news articles. I would like that at a given moment I do not use the news and predict the value until I have other values. ...
0
votes
0answers
8 views

How to deal with possible heterogeneity in predictive performance?

I have a nonlinear regression model that I want to fit to data with Least Absolute Deviation. The model is going to be used for predictive purposes. The data can be divided into two subsets, that ...
0
votes
0answers
3 views

planned comparisons across a continuous gradient

I am analyzing data on stream community before and after an invasive species removal event and have a dataset of dissimilarity values comparing the stream community at each time point in the dataset ...
2
votes
2answers
50 views

Forecasting the number of visitors in each hotel in a city

I am looking for some suggestion on what a good approach would be for the following forecasting problem. Problem statement: There are 100 hotels in a city and I have the monthly data on the total ...
9
votes
7answers
245 views
+50

Avoiding social discrimination in model building

I have questions inspired from the Amazon recent recruitment scandal, where they were accused of discrimination against women in their recruitment process. More info here: Amazon.com Inc's machine-...
1
vote
1answer
30 views

With test accuracy being equal, is it better to have lower training accuracy?

Suppose we train two models on a training set, and then test them both on the training set itself, and on a test set. We have some accuracy metric we're using to evaluate them. Both models score ...
1
vote
0answers
23 views

What is the integral of the False Positive Rate over the False Positive Rate, compared to the AUC?

In machine learning the Area Under the Receiver Operating Characteristic Curve ($AUC$) can be illustrated in a plot of the True Positive Rate ($TPR$) against the False Positive Rate ($FPR$). Formally, ...
0
votes
0answers
29 views

How do I include new features that did not exist before into an existing model?

I have a binary classification model predicting sports result with features covering 10 years worth of matches. However, how would I feed new tracking data that is only limited to the last 3 years. ...
0
votes
0answers
4 views

Predict / recommend products based user historic purchases

I am making a trend based product recommendation based on order data of users. I have users which has a trend of buying item A everyday, item B every third day, item c after 10 days and so on. I ...
1
vote
3answers
89 views

Random Forests performs far less better than Naive Bayes

In one of my Machine Learning courses I have to find the best predictor for this dataset and its binary target "Caesarian". First of all, I tried to improve the datas : there are few features. I did ...
1
vote
1answer
55 views

linear regression predicts lower than expected

I am trying to predict first term GPA for college students based on a number of incoming factors (high school gpa, placement test, year). This isn't the overall model just a simpler one. The first ...
0
votes
0answers
25 views

Forecasting Short Time-Series with other time-series

This is sort of related to a previous question, but now I don't have the requirement of generating customer-level forecasts. I've acquired a set of card customers every month for the last 3 years. I ...
0
votes
1answer
43 views

How to predict values up to two weeks ahead?

I have a data frame that displays transaction numbers in 30-minute blocks since March 2016 to June 2018. It consists of over 40,000 rows of data and looks like this: ...
0
votes
0answers
14 views

Incorporating prior knowledge into feature selection in the setting of multicollinearity?

Background: I'm trying to find the optimal combination of two parameters for finding the first peak meeting some criteria in a signal. The filtering is a bit simplistic, there's a threshold (...
0
votes
0answers
9 views

Estimate expected power on a wind turbine based on other nearby wind turbines

I'm looking for a reliable way to estimate the power that a wind turbine should be producing, based on the power that its neighbours are producing. We use this to identify turbines that are ...
2
votes
2answers
83 views

Highly correlated engineered features any helpful?

Take a car price predictor for an example. If you know the model and year of a car, you can extrapolate facts ("engineer features") about the car. For example: city and highway mpg, number of doors, ...
1
vote
0answers
31 views

Building and interpretation of regression models [closed]

Coming from the world of classification I struggle to understand how to properly build and interpret regression models. Some tutorials only build models on the training data to make conclusions. ...
1
vote
0answers
21 views

With two deep learning models, how do I perform Bayesian Model Averaging for better prediction on a test set?

Given two deep learning models that can predict on a test set, what I want to do is use BMA (Bayesian Model Averaging) to average the models to better predict? What exactly is the procedure for this? ...
1
vote
0answers
32 views

Regression with a high number of 0's for target variable. How do I approach this?

I have a dataset where the probability of an event happening is very low (15%-20%). When the event happens, there's a dollar amount attached to it. The distribution is very right skewed, ranging from -...
0
votes
0answers
8 views

How to assign well-calibrated probabilites of success to multiple competing individuals based on their features

Say I want to predict the probabilities that a large number of students will win a scholarship based on a few metrics (SAT scores, GPA, class rank, etc.). There may be 500 applicants, but only a ...
0
votes
0answers
14 views

model selection for peak findings

I need to create a model, that finds when there will be a significant chance of occurring High peak in the signal. My data has N length of inputs and corresponding N length of outputs. For example, ...
0
votes
0answers
24 views

Should I stack models or extract more features for a tiny, but hard gain in R2?

I heard that stacking models is only worth it doing it in a Kaggle competition as everyone is dealing with the same training data, and due to time limit, feature engineering only helps a little with ...
2
votes
1answer
26 views

How to get model in knn()?

Given I have classified my inputs using R's built-in knn(): ...
1
vote
0answers
29 views

Predictive models to predict sales with r

I would like to find a good model to predict which client will buy my product in 2018. I would like to have opinions on which method can fit my data to predict which client will the product A in 2018. ...
0
votes
0answers
20 views

Can a permutation test be used to evaluate a model?

Let's assume that we have used a given predictive model to generate predictions for an evaluation data set. Now I would need to use these predictions to decide how good is the model. Or, more ...
0
votes
0answers
38 views

How do we know that a model (really) has a predictive power?

Let's assume that we have a trained predictive model and some data set to validate / evaluate the model. We also have a measure of accuracy (mean squared deviation). We apply the given model to a ...
0
votes
1answer
40 views

Is it overfitting if I am using predictions from cross-validation as a level 2 feature for stacking model?

I am learning how to stack models, but I am worried if this is not a practical way to do it. I am using the full dataset and using cross_val_predict to get the ...
0
votes
1answer
37 views

Estimate a probability distribution of target values using features

In my particular problem, I have $$t \in \{1,...N\}$$ time periods, and feature vectors $$x_t \in R^m $$ which I hypothesize predict something about the probability distribution that the targets $...
0
votes
0answers
11 views

Making a more reactive model

I am working on a model to predict quarterly values. I am running into an issue where my back data has extreme trends and my model (I am using a Holt-Winters model) seems to be taking these old values ...
0
votes
0answers
56 views

Is bootstrapping suitable for deriving prediction intervals in models which randomly sample from distributions?

I'm working with a fairly complex predictive model which essentially produces total populations for different groups in future years. Joiners, leavers, and transitions between the groups are modeled ...
0
votes
1answer
51 views

How to determine if one predictive model is statistically significantly better than another one?

I have a data set, two competing predictive models (regressions) and I need to decide which predictive model is better. Let us also assume that I have a measure of accuracy (for example mean squared ...
0
votes
1answer
28 views

data set 'bootstrapping'. Is there a name for this?

First off, I'd like to make clear the term I use 'bootstrapping' is not the statistical technique, but rather the more general and olden phrase of 'pulling yourself up by the bootstraps'. The reason ...
0
votes
0answers
17 views

Variance of error term is nonconstant between observations

I used XGB algorithm to train a model. The task is to train models to predict human personality based on his/her personal photo. We found some significant features when we extracted them by Pearson ...
1
vote
1answer
35 views

During a regression task, I am getting low $R^2$ values, but elementwise difference between test set and prediction values is huge

I am doing a random forest regression on my dataset (which has abut 15 input features and 1 target feature). I am getting a decently low $R^2$ of <1 for both the train and test sets (please do let ...
0
votes
1answer
51 views

When doing Cox regression, what should we take as event duration for censored data, and how should we divide the data into training and testing sets?

I am starting out with survival analysis, and am confused about some things specifically in Cox regression. I have not found a tutorial that explains these things clearly, that's why I'm posting this ...
1
vote
0answers
43 views

Evaluate the conditional variance forecast from a GARCH model

I wanna evaluate a simple GARCH(1,1) model for the conditional variance. Firstly, I understand that the conditional variance is unobserved and that is really the crux of the issue. Out-of-sample, I ...