Stack Overflow
Stack Overflow

Questions tagged [experimental-design]

Experimental design is the methodological study of how experiments should be designed to maximize the scientist's ability to make inferences about the causal effect of an independent variable on a dependent variable.

138 questions
1
vote
1 answer

What type of database for storing ML experiments

So I'm thinking to write some small piece of software, which to run/execute ML experiments on a cluster or arbitrary abstracted executor and then save them such that I can view them in real time efficiently. The executor software will have access…
Alex Botev
  • 1,369
  • 2
  • 19
  • 34
1
vote
0 answers

Confounded columns in R vs Statsmodels - Statsmodels splits values among all aliases

When I have an underconstrained system with confounded columns, lm in R ignores many second- and third-factor interactions (which to me seems correct behaviour), but statsmodels (in Python) splits the value among all confounded columns. Imagine I…
Jean Nassar
  • 555
  • 1
  • 5
  • 13
1
vote
1 answer

How to handle duplicated recommendations in an online experiment for the recommender system

I am about to run an online user experiment to compare different strategies of the recommender system. I will compare 18 strategies and each strategy produces five recommendations, thus I have to ask subjects to evaluate 90 recommendations in…
Benben
  • 1,355
  • 5
  • 18
  • 31
1
vote
1 answer

getting inconsistent output between two methods of reversing a list

I am trying to run a counter balancing program. I cannot understand what Python is doing to get the output I receive. I have 8 variants of 16 maps. Variants are defined by timingConditions * distractionConditions. There are 4 key parts of the…
Doctor David Anderson
  • 274
  • 3
  • 13
1
vote
3 answers

unique values of rows

I often encounter data that looks like this: #create dummy data frame data <- as.data.frame(diag(4)) data[data==0] <- NA data[2,2] <- NA data #V1 V2 V3 V4 #1 1 NA NA NA #2 NA NA NA NA #3 NA NA 1 NA #4 NA NA NA 1 Rows represent participants and…
Flo
  • 1,503
  • 1
  • 18
  • 35
1
vote
2 answers

Calculate D-efficiency of an experimental desgin in R

I have an experimental design. I want to calculate its D-efficiency. I thought R package AlgDesign could help. I found function optFederov which generates the design and - if the user wants - returns its D efficiency. However, I don't want to use…
user2323534
  • 585
  • 1
  • 6
  • 18
1
vote
1 answer

psychopy builder experiment with feedback and multiple if loops

I am trying to use the Coder view to make an experiment that involves feedback and multiple conditional statements. How do you do this though? The task will ultimately involve 4 math problems, and participants will be allowed to have up to 3…
Jennifer
  • 11
  • 2
1
vote
2 answers

R: Permuted-Block Randomization with Varying Block Sizes

Here is a provisional solution to the question I posed earlier. It seems to work, but when I look at the shuffling that ddply does I wonder if it is, in fact, random. Some of the splits seem unfortunate, like BAAAAB in block 4. I had thought about…
Paul
  • 383
  • 4
  • 14
0
votes
0 answers

How do I correctly run regression-adjustment on a ratio metric?

I have some experiment data where I have been using the CUPED methodology outlined in a microsoft paper to adjust the post-assignment data using the pre-assignment data. One of my metrics is a composite metric where I have, for each user, the total…
John Wigg
  • 1
0
votes
0 answers

Weights And Biases Sweep Across Multiple Datasets

TLDR; How can I conduct a sweep in weights and biases across not just hyperparameters, but across model architectures and datasets as well, in a way that respects sensible aggregation and interpretability? I'm new to using weights and biases, and…
Warlax56
  • 1,170
  • 5
  • 30
0
votes
0 answers

What type of design is this experiment? 3x2x2 Factorial design?

I am running an experiment to find out if the pattern of expected and actual guilt people feel when making and not making a purchase is dependent on proximity to their payday (e.g. do people expect/actually feel more guilt about making than not…
user161652
  • 1
0
votes
0 answers

Compiling Data from 2 DoE

I have results from two DoEs, the first one includes 4 factors: | Variable | Levels | | -------- | -------------- | | x1 | 0.9, 2.1, 2.9 | | x2 | 1.64, 2.16 | | x3 | 0.8, 1.15, 1.5 | | x4 | 7, 7.5, 8 | nruns…
Salsa
  • 1
  • 1
0
votes
0 answers

How to visualize experimental data for a single independent variable that depends on 5 discrete variables?

I'm trying to visualize the efficiency of a coaxial UAV propeller test stand with variable coaxial spacing, variable pitch, and variable motor speed. My independent variables are coaxial separation, front motor speed, back motor speed, front motor…
DanielB
  • 1
  • 1
0
votes
0 answers

Does it make sense to do the power analysis if one of the sample proportions is 0%?

Can we have any sample size if one of the sample proportions is 0. I mean even the slightest difference detected between the two groups would be because of the treatment in this case, and hence I do not see any reason why the power analysis should…
learner101
  • 63
  • 2
  • 5
0
votes
0 answers

Specifying quadratic models for D-optimal designs in AlgDesign - quad function not found

I need to optimize a production process by way of adjusting a total of 6 factors, all of them numerical variables with 3 levels. I expect the responses to be related to both the main effects and their quadratic terms, and I expect some of the second…
Enrico P
  • 1
1 2 3
…
9 10