# Strategy for improving performance of 3D convolutional GAN

Data Science Asked by BBirdsell on December 22, 2020

Others working with neural nets and GAN’s might find this question interesting.

Background:

I’ve been working with data from Berkeleys PEER Ground Motion Database to generate new novel seismic traces. (Real traces are rendered above.) Coming from a background in engineering my first attempt involved decomposing the traces into their {X,Y,Z} components, however the results were less than satisfying. The repeated mode collapse output can be seen below. There might be ways to fix this with more time and resources but I thought I would try another approach.

Forward:

I still have a bit of time to work on the data, and was looking to solicit methods for turning this {X,Y,Z} point data into something more digestible by a 3D convolutional network. I have given a small sample of one of the traces of 3 features each over time. Each trace is scaled {-1,1) across all axis and interpolated to 4000 steps. (Nearly all of the 788 traces in the training data are over 4000 steps and were down sampled.)

{{0.095746,-0.301555,-0.407207},{0.0955857,-0.301693,-0.407324},{0.0953887,-0.301861,-0.407461},{0.095148,-0.302067,-0.407609},{0.0948748,-0.302296,-0.407768},{0.0945636,-0.302547,-0.407985},{0.0942212,-0.3028,-0.40827},{0.0938457,-0.30301,-0.408664},{0.0934454,-0.303154,-0.409237},{0.0929839,-0.303259,-0.410011},{0.0924333,-0.303451,-0.410692},{0.0917118,-0.3039,-0.41102},{0.0907978,-0.304671,-0.411107},{0.0897068,-0.306049,-0.410927},{0.0886644,-0.308036,-0.410422},{0.0878034,-0.310481,-0.409685},{0.087091,-0.313055,-0.409298},{0.0863677,-0.315387,-0.409476},{0.0854884,-0.317162,-0.409922},{0.0845426,-0.31811,-0.410777}}


My knowledge of the subject matter suggests I need to transform this data into some type of array with places a 1 if there is a trace point there and 0 if void. Is that correct?

Before jumping in and creating another big branch of code, I wanted to confirm some of my assumptions moving forward. It seems like a rather computationally expensive method to divide these real valued 2x2x2 volumes and essentially count how many points are in each volume to recreate the geometry. Are there any precedence for this that I can follow? Is there a way to do this directly with a 3D convolution?

## Related Questions

### Appearance-based hashing for similarity detection for picking the 100 most distinct images out 500 images

0  Asked on April 1, 2021

### Trained BERT models perform unpredictably on test set

1  Asked on April 1, 2021 by peterpaul

### How to get a KNN model (using quantiles to scale variables due to non-normal distributed data) to be better suited for non-extreme values in the data?

0  Asked on April 1, 2021 by sh_student

### Deep learning and label noise. Best practices for the real world

0  Asked on April 1, 2021 by ajeje

### Linear and logistic regression output with same neural network

0  Asked on March 31, 2021 by tz39r

### Same validation accuracy, different train accuracy for two neural networks models

2  Asked on March 31, 2021 by sneaky_lobster

### What methods are used to determine acceptable loss levels for autoencoders?

1  Asked on March 31, 2021 by hrokr

### What are some of the best practices for sharing data and models with colleagues?

6  Asked on March 31, 2021 by asampat3090

### How to predict auction winning price without knowing the winning price when you lose the auction?

0  Asked on March 31, 2021 by vance

### Why does horizontal lines in plt.plot(feature, ‘.’) mean that the data have been properly shuffled?

1  Asked on March 31, 2021

### Preferred approaches for imbalanced data

1  Asked on March 31, 2021

### Can we use feature selection and dimensionality reduction together?

1  Asked on March 31, 2021 by dinz_n

### Calculation of PCA

0  Asked on March 31, 2021

### How to get weightage of each factor for a prediction in a neural network classifier?

1  Asked on March 31, 2021 by ssy

### Question about Similarity vs Dissimilarity Matrix

2  Asked on March 30, 2021 by thomas-formal

### Is it possible to modify or replace the basic network of YOLO?

0  Asked on March 30, 2021 by colinguolin

### KMeans clusterization on documents

2  Asked on March 30, 2021