Questions tagged [synthetic-data]

The tag has no usage guidance.

Filter by
Sorted by
Tagged with
-1
votes
0answers
18 views

Synthetic Data - Can it be useful? [closed]

Anyone here built and trained models using synthetic data for real world projects? Curious to get thoughts on how effective people have found it
1
vote
0answers
27 views

Generate data that matches a frequency distribution while preserving the original spatial structure

I am dealing with a 3D array containing values representing the "importance" of each voxel. For my analysis, I would like to synthesize n new arrays from my original array to have a ...
0
votes
0answers
32 views

How can I generate synthetic data that fits with population statistics from multiple two-way tables?

I have a set of statistical tables (from the Northern Ireland House Condition Survey 2016), each containing counts of homes in each category. Central Heating type by Home Age Band Dwelling Type by ...
0
votes
0answers
14 views

Is creating artificial class imbalance in synthetic training data a good way to tackle hard cases in classification?

I have a problem where I need to classify something around 50 different classes. Some of the classes are very similar to one another and the algorithm tends to confuse them. However, I can create a ...
1
vote
0answers
12 views

Learning Distribution of Data [duplicate]

Sometimes it's important to generate data due to data imbalance issues. I heard that data augmentation by leaning distribution of data is a hot topic now. Could you please give me some resources and ...
0
votes
0answers
9 views

Bayesian network: graph synthesis & data sampling

Input (What I have): some Bayesian networks (both graph structure and conditional probability distribution (cpd)) and corresponding categorical datasets (e.g. bnlearn repo). Output (What I want): ...
0
votes
2answers
23 views

How to generate synthetic data from a balanced dataset?

Let say I have a balanced dataset that has a small training sample size (lack of data). How do I increase the training sample size by generating synthetic data based on the original data? I believe ...
0
votes
1answer
56 views

What can be a good way to generate data(tabular) from statistical facts and probability of data?

For Example, If I have facts saying that: 50% of humans are male 30% of males are Indians 70% of Indians are brown average age of Indians are 27 30% of indian females are working 80% of no-indian ...
0
votes
1answer
27 views

Simulated data for statistical framework testing

I want to generate two sets of simulated data (numeric) for statistical testing. Is it possible to generate datasets with predefined RMSE or accuracy?
0
votes
0answers
24 views

Design synthetic data for regression problem that give access to the true risk value

Consider the following linear regression model: $Y = X \beta + \epsilon$ Let $\hat{\beta}$ be the OLS estimator. It verifies the normal equation i.e $\langle X, X\hat{\beta} - Y\rangle = 0 \iff \hat{\...
0
votes
0answers
20 views

Synthetic or simulated data

I need to develop a set of simulated or synthetic (numeric) datasets to evaluate a data fusion framework I have proposed. I already have a real-world data set but I need to test the framework for some ...
2
votes
1answer
72 views

Generating samples on an exponential distribution

I am trying to generate a synthetic earthquake database where the number of events ($N$) with magnitude ($M$) in the range $[M, M+\delta_M]$ follows: $\log_{10}(N) = a - bM$ where $a$ and $b$ are ...
0
votes
0answers
8 views

Should we trust the quality of synthetic data generating from SMOTE or ADASYN?

Real datasets always suffer from imbalance, that why technique like smote and adasyn are invented to generate synthetic data to avoid oversampling or under sampling. But does these synthetic data and ...
3
votes
1answer
89 views

Does it make sense to use the KL-divergence between joint distributions of synthetic and real data, as a evaluation metric?

The KL-divergence is defined as: $D_{KL}(p(x_1)∥q(x_1))=\sum p(x_1)\, \log \Big( \dfrac{p(x_1)}{q(x_1)} \Big)$ I consider the Kullback-Leibler (KL) divergence as a performance metric for data ...
0
votes
0answers
23 views

Constrains on the coefficients of SARIMA

I am trying to generate synthetic time-series through SARIMA random process by defining the model coefficients manually. Could any one help me how to generate coefficients? What are the constraints on ...
3
votes
0answers
15 views

Correlation structures in synthetic experiments

Sorry if this question is too broad. I've noticed that in various statistical articles involving synthetic linear regression experiments, the following similar data generating processes may be ...
0
votes
0answers
23 views

How to use Synth for comparative case studies?

I am trying to use synth for a comparative case study. I went through the documentation of the same, read the original research paper but I am still lost on how to actually implement it? which feature ...
0
votes
0answers
10 views

Create synthetic data based on general description of molecular characteristic for modeling

I have a modeling task that classifies cancer subtype based on several molecular features. However, I do not have raw data of tumor sample to extract these features. All I have are several published ...
0
votes
0answers
79 views

Generate more data for a small dataset

I have been working on a dataset which has 14 attributes and 303 rows(instances) along with the binary labels. I want to generate more data so that I could train my neural networks so that I could ...
1
vote
0answers
42 views

Regression on unevenly distributed high dimensional dataset

I have a very high dimensional (20K+ hand engineered features) biological dataset to predict a single continuous output variable (such as a mental state exam scores for a dementia patient). The output ...
0
votes
1answer
49 views

how to synthesize a virtual control arm from multiple historical trials with different treatment in their control arms

We want to synthesize a virtual control arm for a single arm study, and we found two trials with data available that targeting the same population. One trial used Placebo+Standard of care in the ...
23
votes
10answers
6k views

Best term for made-up data?

I'm writing an example and have made up some data. I want it to be clear to the reader this is not real data, but I also don't want to give the impression of malice, since it just serves as an ...
0
votes
1answer
67 views

Generation of synthetic data for Hierarchical clustering

I wanted to test various hierarchical clustering algorithms to check which algorithm performs best. For this, I was considering simulating some ground truth. Is the possible to generate a correlation ...
0
votes
1answer
83 views

Combining treatment units for synthetic control method

Cross-posting this from the Stata forum to increase visibility - but let me know if that's an issue! I am thinking about using the synthetic control method to evaluate an intervention that took place ...
0
votes
0answers
50 views

Using Synthetic Data Based on Real Data for Classification

The goal is to classify three different cell types based on certain features (e.g. area, shape tensor etc.). However the amount of labelled training data I have is very small. Therefore, it was ...
0
votes
1answer
33 views

Synthetic Control Method with Correlated Variables

I am interested in using the Synthetic Control Method to measure the effects of a certain policy. I am afraid that the synthetic control will not be accurate because the underlying variables that make ...
1
vote
0answers
70 views

Creating synthetic data from existing data

I have data with 8 features and 900 rows each. Those features are dependent on each other in some way. I want to conduct machine learning, but the data size is small. I was suggested to generate ...
2
votes
0answers
128 views

Generating Correlated Random Variables following Weibull Distribution

I am working on analyzing power generation from geographically distributed wind farms. I have wind speed data that have been collected at multiple locations, but the measurement periods for different ...
0
votes
1answer
25 views

Sample whole number from distribution with average less than 1

I am trying to create some simulated data, where for each participant, the average number of events that occur in a given week can be any positive number, but usually in the range of 0 to 5. Trouble ...
0
votes
0answers
110 views

Handling imbalanced data for classification [duplicate]

What are the best ways to deal with imbalanced datasets for classifying whether or not individuals pay their tuition? The data is 75% positive class (paid) and 25% negative (unpaid). Some approaches I ...
0
votes
0answers
37 views

Data Synthesis based on Deep Learning

Is there any open source tool available to synthetically generate a new dataset with the same statistics than the original one? The objective is to create a new tabular dataset that is private but ...
0
votes
0answers
2k views

“Add White Gaussian Noise with SNR” vs. “Add 5% Gaussian Noise”

I have a noise-free dataset which is a vector of numbers, $\mathbf{d}$, with length $N$. I want to "add noise" to this data. My understanding is that there are two ways to do this. 1) Add some ...
12
votes
1answer
383 views

Creating an Imbalanced Dataset

I would like to have my trained model tested on an imbalanced dataset. Is there any algorithms available to generate synthetic data from a balanced labelled dataset (spam/non-spam)?
1
vote
0answers
103 views

Methods for generating regression data whose target follows an arbitrary distribution

I'd like to generate a regression dataset (independent vars that combine somehow into a target or dependent variable). Generating such a dataset is, in general, easy: we can e.g. pick some $x_i$ at ...
3
votes
0answers
46 views

Is it possible to use a time series to construct a synthetic control variable?

Imagine I want to reconstruct a counterfactual euro-dollar exchange rate in 2010 using a synthetic control variable to assess the impact of some policy. Could I use, for example, the exchange rates ...
-1
votes
1answer
294 views

Generalizing “Causal Impact” synthetic controls, to multiple outcomes

Does anybody know a way to generalize the use of the Causal Impact google R package to multiple outcome time series? Say I ran a time series experiment and was able to set up multiple test outcome ...
1
vote
1answer
347 views

Generating synthetic data for robust regression?

I was running some test to see the benefits of using robust regression over ols. For that purposes, I created synthetic data in the following manner, A is the ...
0
votes
2answers
150 views

Generating a synthetic dataset from a latent class model

In latent class analysis, the estimated model paramters are class sizes and item-response probablities in each class. Based on these basic parameters, is it possible to generate a synthetic sample ...
1
vote
0answers
49 views

How to evaluate whether the mean of the synthetic time series is statistically equal to the historical mean

I am testing a stochastic model that generates synthetic hydrologic time series (five-year series, 60(month) data points). I want to evaluate whether these synthetic time series have the same ...
1
vote
0answers
45 views

Using fake datasets to optimize a model

I have a machine learning application I'm working on that will make predictions based on data about retail stores. Problem is, any data collected by BI regarding the performance of a given retail ...
1
vote
1answer
27 views

Generating a matrix so that correlations amongst columns are close to pre-defined values

I want to generate a matrix of values such that the correlations (e.g., Pearsons correlations) are close to a pre-defined set of values (e.g., ...
0
votes
1answer
103 views

Over-fitting because of Artificial Synthesis of Data

In the process of artificial data synthesis in Machine Learning applications, when we introduce noise to generate new data, is it possible that overfitting can occur because we are essentially ...
5
votes
0answers
94 views

Is it possible to determine if a dataset is real or randomly generated?

I've been tasked with developing regression and classification models for time series data. For each observation I have a continuous target for regression and a discrete target for classification. I'...
1
vote
1answer
194 views

When to use synthetic data and when to use regularization parameters to avoid the over fitting and which is better?

Can anyone explain me when to consider generating the synthetic data or when to consider regularization parameters to reduce the error so the machine learning model will not overfit
0
votes
1answer
47 views

Synthetic Control when leaving treatment

I'm trying to do a comparative analysis using synthetic control to compare a country that went from a dictatorship to democracy to it's dictatorial synthetic twin. As I understand it, synthetic ...
0
votes
0answers
26 views

Validating decomposition of Synthetic Tensor generated from Unevenly Sampled Tensor

I have a 3-way tensor generated from 7 experiments, with each experiment being matricized and becoming a frontal slice of the tensor (thus mode-3 is of length 7). The data is generated from ...
1
vote
1answer
92 views

How to build Predictive models with insufficient historical/performance data

I'm building a auto loan probability of default model where the loan term could be 3 to 7 years and hence default can happen anytime in that interval. But we are a start-up and have only 3 years of ...
3
votes
1answer
1k views

Create synthetic data with a given intraclass correlation coefficient (ICC)?

I want to generate some synthetic data with $I$ observations across $J$ clusters. Additionally, I want the intraclass correlation coefficient ($ICC$) to be an input of my data generation process. So, ...
0
votes
2answers
605 views

Generate non-random data that follows a given correlation matrix

Problem: I have to generate in R a synthetic dataset which contains three different variables that didn't follow any known distribution. What I did until now: Grouping historical values of each ...
3
votes
0answers
412 views

How to generate synthetic data with a given $R_{x,y}^2$

I would like to generate some data with the following relationships: $ y = x\beta + T\delta + \varepsilon $ $ R_{x,y}^2 = a $, where $a$ is a number that I can choose when generating the data $ \...