Title: Does Active Learning Reduce Human Coding?: A Systematic Comparison of a Neural Network with nCoder
In quantitative ethnography (QE) studies which often involve large da-tasets that cannot be entirely hand-coded by human raters, researchers have used supervised machine learning approaches to develop automated classi-fiers. However, QE researchers are rightly concerned with the amount of human coding that may be required to develop classifiers that achieve the high levels of accuracy that QE studies typically require. In this study, we compare a neural network, a powerful traditional supervised learning ap-proach, with nCoder, an active learning technique commonly used in QE studies, to determine which technique requires the least human coding to produce a sufficiently accurate classifier. To do this, we constructed multi-ple training sets from a large dataset used in prior QE studies and designed a Monte Carlo simulation to test the performance of the two techniques sys-tematically. Our results show that nCoder can achieve high predictive accu-racy with significantly less human-coded data than a neural network. more »« less
Cai, Z.; Marquart, C.; Shaffer, D.(
, Proceedings of the 15th International Conference on Educational Data Mining)
Mitrovic, A.; Bosch, N.
(Ed.)
Regular expression (regex) coding has advantages for text analysis. Humans are often able to quickly construct intelligible coding rules with high precision. That is, researchers can identify words and word patterns that correctly classify examples of a particular concept. And, it is often easy to identify false positives and improve the regex classifier so that the positive items are accurately captured. However, ensuring that a regex list is complete is a bigger challenge, because the concepts to be identified in data are often sparsely distributed, which makes it difficult to identify examples of \textit{false negatives}. For this reason, regex-based classifiers suffer by having low recall. That is, it often misses items that should be classified as positive. In this paper, we provide a neural network solution to this problem by identifying a \textit{negative reversion set}, in which false negative items occur much more frequently than in the data set as a whole. Thus, the regex classifier can be more quickly improved by adding missing regexes based on the false negatives found from the negative reversion set. This study used an existing data set collected from a simulation-based learning environment for which researchers had previously defined six codes and developed classifiers with validated regex lists. We randomly constructed incomplete (partial) regex lists and used neural network models to identify negative reversion sets in which the frequency of false negatives increased from a range of 3\\%-8\\% in the full data set to a range of 12\\%-52\\% in the negative reversion set. Based on this finding, we propose an interactive coding mechanism in which human-developed regex classifiers provide input for training machine learning algorithms and machine learning algorithms ``smartly" select highly suspected false negative items for human to more quickly develop regex classifiers.
The use of metacognition is critical to learning, especially in fields such as engineering that involve problem‐solving and difficult conceptual material. Due to limitations with current methodological approaches, new methods are needed to investigate engineering students' metacognitive engagement in learning situations that are self‐directed, such as study groups.
Purpose
Our purpose was to develop an approach to investigate the metacognitive engagement of undergraduate engineering students in self‐directed learning environments. The Naturalistic Observations of Metacognition in Engineering (NOME) Observational Protocol and Coding Strategy is a qualitative data collection method that allows researchers to observe the behaviors of students who are studying in groups to determine the student's engagement in different metacognitive practices. The NOME is intended to be used by researchers interested in studying online metacognitive behaviors without the direct interference of a methodological approach.
Design/Method
We observed three study groups where students were working on an engineering problem‐solving homework assignment. Using a taxonomic definition of metacognition, we coded episodes of observation transcripts to identify behaviors that represented key definitions in the taxonomy.
Results
We combined subcodes and descriptions of behaviors with key definitions to develop a coding strategy useful for future observational studies. Evidence of intercoder agreement and agreement in unitizing indicates that the coding strategy can reliably be used by multiple trained coders to identify metacognitive engagement.
Conclusions
The reliability evidence shows that the NOME may be a useful tool for researchers in engineering education interested in studying the metacognitive habits of engineering students in self‐directed study.
Jean-Paul, S.; Elseify, T.; Obeid, I.; Picone, J.(
, IEEE Signal Processing in Medicine and Biology Symposium (SPMB))
Obeid, I.; Selesnik, I.; Picone, J.
(Ed.)
The Neuronix high-performance computing cluster allows us to conduct extensive machine learning experiments on big data [1]. This heterogeneous cluster uses innovative scheduling technology, Slurm [2], that manages a network of CPUs and graphics processing units (GPUs). The GPU farm consists of a variety of processors ranging from low-end consumer grade devices such as the Nvidia GTX 970 to higher-end devices such as the GeForce RTX 2080. These GPUs are essential to our research since they allow extremely compute-intensive deep learning tasks to be executed on massive data resources such as the TUH EEG Corpus [2]. We use TensorFlow [3] as the core machine learning library for our deep learning systems, and routinely employ multiple GPUs to accelerate the training process.
Reproducible results are essential to machine learning research. Reproducibility in this context means the ability to replicate an existing experiment – performance metrics such as error rates should be identical and floating-point calculations should match closely. Three examples of ways we typically expect an experiment to be replicable are: (1) The same job run on the same processor should produce the same results each time it is run. (2) A job run on a CPU and GPU should produce identical results. (3) A job should produce comparable results if the data is presented in a different order. System optimization requires an ability to directly compare error rates for algorithms evaluated under comparable operating conditions. However, it is a difficult task to exactly reproduce the results for large, complex deep learning systems that often require more than a trillion calculations per experiment [5]. This is a fairly well-known issue and one we will explore in this poster.
Researchers must be able to replicate results on a specific data set to establish the integrity of an implementation. They can then use that implementation as a baseline for comparison purposes. A lack of reproducibility makes it very difficult to debug algorithms and validate changes to the system. Equally important, since many results in deep learning research are dependent on the order in which the system is exposed to the data, the specific processors used, and even the order in which those processors are accessed, it becomes a challenging problem to compare two algorithms since each system must be individually optimized for a specific data set or processor. This is extremely time-consuming for algorithm research in which a single run often taxes a computing environment to its limits. Well-known techniques such as cross-validation [5,6] can be used to mitigate these effects, but this is also computationally expensive.
These issues are further compounded by the fact that most deep learning algorithms are susceptible to the way computational noise propagates through the system. GPUs are particularly notorious for this because, in a clustered environment, it becomes more difficult to control which processors are used at various points in time. Another equally frustrating issue is that upgrades to the deep learning package, such as the transition from TensorFlow v1.9 to v1.13, can also result in large fluctuations in error rates when re-running the same experiment. Since TensorFlow is constantly updating functions to support GPU use, maintaining an historical archive of experimental results that can be used to calibrate algorithm research is quite a challenge. This makes it very difficult to optimize the system or select the best configurations.
The overall impact of all of these issues described above is significant as error rates can fluctuate by as much as 25% due to these types of computational issues. Cross-validation is one technique used to mitigate this, but that is expensive since you need to do multiple runs over the data, which further taxes a computing infrastructure already running at max capacity.
GPUs are preferred when training a large network since these systems train at least two orders of magnitude faster than CPUs [7]. Large-scale experiments are simply not feasible without using GPUs. However, there is a tradeoff to gain this performance. Since all our GPUs use the NVIDIA CUDA® Deep Neural Network library (cuDNN) [8], a GPU-accelerated library of primitives for deep neural networks, it adds an element of randomness into the experiment. When a GPU is used to train a network in TensorFlow, it automatically searches for a cuDNN implementation. NVIDIA’s cuDNN implementation provides algorithms that increase the performance and help the model train quicker, but they are non-deterministic algorithms [9,10]. Since our networks have many complex layers, there is no easy way to avoid this randomness. Instead of comparing each epoch, we compare the average performance of the experiment because it gives us a hint of how our model is performing per experiment, and if the changes we make are efficient.
In this poster, we will discuss a variety of issues related to reproducibility and introduce ways we mitigate these effects. For example, TensorFlow uses a random number generator (RNG) which is not seeded by default. TensorFlow determines the initialization point and how certain functions execute using the RNG. The solution for this is seeding all the necessary components before training the model. This forces TensorFlow to use the same initialization point and sets how certain layers work (e.g., dropout layers). However, seeding all the RNGs will not guarantee a controlled experiment. Other variables can affect the outcome of the experiment such as training using GPUs, allowing multi-threading on CPUs, using certain layers, etc.
To mitigate our problems with reproducibility, we first make sure that the data is processed in the same order during training. Therefore, we save the data from the last experiment and to make sure the newer experiment follows the same order. If we allow the data to be shuffled, it can affect the performance due to how the model was exposed to the data. We also specify the float data type to be 32-bit since Python defaults to 64-bit. We try to avoid using 64-bit precision because the numbers produced by a GPU can vary significantly depending on the GPU architecture [11-13]. Controlling precision somewhat reduces differences due to computational noise even though technically it increases the amount of computational noise.
We are currently developing more advanced techniques for preserving the efficiency of our training process while also maintaining the ability to reproduce models. In our poster presentation we will demonstrate these issues using some novel visualization tools, present several examples of the extent to which these issues influence research results on electroencephalography (EEG) and digital pathology experiments and introduce new ways to manage such computational issues.
Generative Adversarial Networks (GANs) recently demonstrated a great opportunity toward unsupervised learning with the intention to mitigate the massive human efforts on data labeling in supervised learning algorithms. GAN combines a generative model and a discriminative model to oppose each other in an adversarial situation to refine their abilities. Existing nonvolatile memory based machine learning accelerators, however, could not support the computational needs required by GAN training. Specifically, the generator utilizes a new operator, called transposed convolution, which introduces significant resource underutilization when executed on conventional neural network accelerators as it inserts massive zeros in its input before a convolution operation. In this work, we propose a novel computational deformation technique that synergistically optimizes the forward and backward functions in transposed convolution to eliminate the large resource underutilization. In addition, we present dedicated control units - a dataflow mapper and an operation scheduler, to support the proposed execution model with high parallelism and low energy consumption. ZARA is implemented with commodity ReRAM chips, and experimental results show that our design can improve GAN’s training performance by averagely 1.6x~23x over CMOS-based GAN accelerators. Compared to state-of-the-art ReRAM-based accelerator designs, ZARA also provides 1.15x~2.1x performance improvement.
In this paper, we present a novel Deformable Neural Articulations Network (DNA-Net), which is a template- free learning-based method for dynamic 3D human reconstruction from a single RGB-D sequence. Our proposed DNA-Net includes a Neural Articulation Prediction Net- work (NAP-Net), which is capable of representing non-rigid motions of a human by learning to predict a set of articulated bones to follow movements of the human in the in- put sequence. Moreover, DNA-Net also include Signed Distance Field Network (SDF-Net) and Appearance Network (Color-Net), which take advantage of the powerful neural implicit functions in modeling 3D geometries and appear- ance. Finally, to avoid the reliance on external optical flow estimators to obtain deformation cues like previous related works, we propose a novel training loss, namely Easy-to- Hard Geometric-based, which is a simple strategy that inherits the merits of Chamfer distance to achieve good de- formation guidance while still avoiding its limitation of lo- cal mismatches sensitivity. DNA-Net is trained end-to-end in a self-supervised manner directly on the input sequence to obtain 3D reconstructions of the input objects. Quantitative results on videos of DeepDeform dataset show that DNA-Net outperforms related state-of-the-art methods with an adequate gaps, qualitative results additionally prove that our method can reconstruct human shapes with high fidelity and details.
Choi, J., Ruis, A. R., Cai, Z., Eagan, B. R., and Shaffer, D. W. Does Active Learning Reduce Human Coding?: A Systematic Comparison of a Neural Network with nCoder. Retrieved from https://par.nsf.gov/biblio/10354410. Advances in Quantitative Ethnography: Fourth International Conference, International Conference on Quantitative Ethnography 2022 .
Choi, J., Ruis, A. R., Cai, Z., Eagan, B. R., & Shaffer, D. W. Does Active Learning Reduce Human Coding?: A Systematic Comparison of a Neural Network with nCoder. Advances in Quantitative Ethnography: Fourth International Conference, International Conference on Quantitative Ethnography 2022, (). Retrieved from https://par.nsf.gov/biblio/10354410.
Choi, J., Ruis, A. R., Cai, Z., Eagan, B. R., and Shaffer, D. W.
"Does Active Learning Reduce Human Coding?: A Systematic Comparison of a Neural Network with nCoder". Advances in Quantitative Ethnography: Fourth International Conference, International Conference on Quantitative Ethnography 2022 (). Country unknown/Code not available. https://par.nsf.gov/biblio/10354410.
@article{osti_10354410,
place = {Country unknown/Code not available},
title = {Does Active Learning Reduce Human Coding?: A Systematic Comparison of a Neural Network with nCoder},
url = {https://par.nsf.gov/biblio/10354410},
abstractNote = {In quantitative ethnography (QE) studies which often involve large da-tasets that cannot be entirely hand-coded by human raters, researchers have used supervised machine learning approaches to develop automated classi-fiers. However, QE researchers are rightly concerned with the amount of human coding that may be required to develop classifiers that achieve the high levels of accuracy that QE studies typically require. In this study, we compare a neural network, a powerful traditional supervised learning ap-proach, with nCoder, an active learning technique commonly used in QE studies, to determine which technique requires the least human coding to produce a sufficiently accurate classifier. To do this, we constructed multi-ple training sets from a large dataset used in prior QE studies and designed a Monte Carlo simulation to test the performance of the two techniques sys-tematically. Our results show that nCoder can achieve high predictive accu-racy with significantly less human-coded data than a neural network.},
journal = {Advances in Quantitative Ethnography: Fourth International Conference, International Conference on Quantitative Ethnography 2022},
author = {Choi, J. and Ruis, A. R. and Cai, Z. and Eagan, B. R. and Shaffer, D. W.},
editor = {Barany, A. and Damsa, C.}
}
Warning: Leaving National Science Foundation Website
You are now leaving the National Science Foundation website to go to a non-government website.
Website:
NSF takes no responsibility for and exercises no control over the views expressed or the accuracy of
the information contained on this site. Also be aware that NSF's privacy policy does not apply to this site.
