Data Science Stack Exchange
Data Science Stack Exchange

Questions tagged [inference]

33 questions
6
votes
1 answer

How to correctly perform link prediction inference on a new, unseen graph?"

I'm working on an industrial AI use case where I train a Graph Neural Network (GCN) for link prediction — specifically, to predict successor tasks in project planning graphs (e.g., for construction or maintenance workflows). For example we have this…
lili
  • 311
  • 6
4
votes
2 answers

how do I calculate p-value?

I need to calculate statistical significance of difference between two time series, each with 4500 terms. My null hypothesis is that $H_0: \mu=\mu_0$. How can I calculate p-value? Is Z-statistic useful for p value calculation? How to get p-value…
nick_name
  • 205
  • 2
  • 8
3
votes
0 answers

How do I make inference about test metrics for entire population from sample metrics?

Generally we calculate specific metrics for ML models on a test set (and we try to make that test set representative). I'm not clear on how to make inference about the same metrics for the population that the test set is representing - i.e., say I…
Shirish
  • 299
  • 1
  • 10
3
votes
1 answer

Which model is the best suitable for generating edges?

I'm trying to develop a model who'd be able to generate dependencies between industrial tasks. In order to do that, i went for the GNN solution : i have nodes = tasks, dependencies = edges, and have multiple graphs (graphs = a list of…
lili
  • 311
  • 6
2
votes
1 answer

How train - test split works for Graph Neural Networks

I have recently started studying GNN's. I have covered GCN and GraphSage so far. But I am confused regarding the process when testing occurs. Now suppose in the graph above I am using the nodes as train and test set as shown in the figure. Suppose…
Sourajit
  • 123
  • 1
  • 4
2
votes
2 answers

p-value and effect size

Is it correct to say that the lower the p-value is the higher is the difference between the two means of the two groups in the t-test? For example, if I apply the t-test between two groups of measurements A and B and then to two groups of…
LearningAlgorithm
  • 114
  • 2
  • 18
2
votes
0 answers

Maximum number of classes YOLO net can recognize on mobile

I'm trying to make a mobile app on image recognition(Computer Vision Application) . Does anyone know whether modern day smartphones have enough processing power/memory to recognize, say about 1 million classes from their real-time camera feed (30…
R. Rudinthkin
  • 21
  • 1
2
votes
1 answer

Exogeneous, Endogenous Variables in Structural Causal Model

https://en.wikipedia.org/wiki/Causal_model#Definition Wikpedia defines causal models as: an ordered triple $\langle U, V, E\rangle$, where $U$ is a set of exogenous variables whose values are determined by factors outside the model; $V$ is a set of…
Eisen
  • 311
  • 2
  • 6
2
votes
2 answers

Causal Network Graph

I've been reading this blog (https://deepmind.com/blog/article/Causal_Bayesian_Networks) and am just getting into Causal inference. I have a question regarding causal network graphs. Let's say given data, how exactly are causal network graphs…
Eisen
  • 311
  • 2
  • 6
1
vote
1 answer

In keras seq2seq model, what is the difference between `model.predict()` and the inference model?

I am looking into seq2seq model in keras, for example, this blog post from keras or this. All the examples I have seen have some inference model, that depicts the original model. That inference model is then used to make the predictions. My question…
BlueMango
  • 113
  • 3
1
vote
2 answers

Combining CNNs for image classification

I would like to take the output of an intermediate layer of a CNN (layer G) and feed it to an intermediate layer of a wider CNN (layer H) to complete the inference. Challenge: The two layers G, H have different dimensions and thus it can't be done…
Andrew
  • 63
  • 5
1
vote
0 answers

ML models that train on graphs but infer without any edges (edge prediction task)

I'm exploring a machine learning research direction and I'm looking for ideas or pointers to existing models/projects that fit the following setup: The model is trained on graphs with edge information (e.g., node features + edges). At inference…
lili
  • 311
  • 6
1
vote
0 answers

Different inference times by factor 10 for two models with similar architecture and training data

I am currently trying to replicate or come close to the results achieved in the Multispider paper, which is a multilingual Text-to-SQL benchmark. I have downloaded the model used by the authors of the paper to create their results. This takes 9/10…
Uwe
  • 11
  • 1
1
vote
1 answer

How to Use Multiple Adapters with a Pretrained Model in Hugging Face Transformers for Inference?

I have a pretrained Llama-2 model in the models_hf directory and two fine-tuned adapters: a summarization adapter in ./tmp/llama-output and a chat adapter in ./tmp/chat_adapter. The details of the code are in another question. For inference, I'd…
Aun Zaidi
  • 133
  • 1
  • 6
1
vote
0 answers

What is the best way to combine cross-validation and bootstrapping for one application?

We intend to model data with non-parametric covariate splines and we would like to understand the uncertainty of the parameter estimates/response estimates. Currently, we use cross-validation to model the optimal smoothness of our spline models…
Stan Tendijck
  • 111
  • 1
1
2 3