More Like this

1. Issues in the Reproducibility of Deep Learning Results

   https://doi.org/10.1109/SPMB47826.2019.9037840  

   Jean-Paul, S. ; Elseify, T. ; Obeid, I. ; Picone, J. ( December 2019 , 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. 

   more » « less
2. The Deep Learning Epilepsy Detection Challenge: Design, Implementation, and Test of a New Crowd-Sourced AI Challenge Ecosystem

   Kiral, Isabell ; Roy, Subhrajit ; Mummert, Todd ; Braz, Alan ; Tsay, Jason ; Tang, Jianbin ; Asif, Umar ; Schaffter, Thomas ; Mehmet, Eren ; Picone, Joseph ; et al ( April 2019 , Challenges in Machine Learning Competitions for All (CiML))

   null (Ed.)

   The DeepLearningEpilepsyDetectionChallenge: design, implementation, andtestofanewcrowd-sourced AIchallengeecosystem Isabell Kiral*, Subhrajit Roy*, Todd Mummert*, Alan Braz*, Jason Tsay, Jianbin Tang, Umar Asif, Thomas Schaffter, Eren Mehmet, The IBM Epilepsy Consortium◊ , Joseph Picone, Iyad Obeid, Bruno De Assis Marques, Stefan Maetschke, Rania Khalaf†, Michal Rosen-Zvi† , Gustavo Stolovitzky† , Mahtab Mirmomeni† , Stefan Harrer† * These authors contributed equally to this work † Corresponding authors: rkhalaf@us.ibm.com, rosen@il.ibm.com, gustavo@us.ibm.com, mahtabm@au1.ibm.com, sharrer@au.ibm.com ◊ Members of the IBM Epilepsy Consortium are listed in the Acknowledgements section J. Picone and I. Obeid are with Temple University, USA. T. Schaffter is with Sage Bionetworks, USA. E. Mehmet is with the University of Illinois at Urbana-Champaign, USA. All other authors are with IBM Research in USA, Israel and Australia. Introduction This decade has seen an ever-growing number of scientific fields benefitting from the advances in machine learning technology and tooling. More recently, this trend reached the medical domain, with applications reaching from cancer diagnosis [1] to the development of brain-machine-interfaces [2]. While Kaggle has pioneered the crowd-sourcing of machine learning challenges to incentivise data scientists from around the world to advance algorithm and model design, the increasing complexity of problem statements demands of participants to be expert data scientists, deeply knowledgeable in at least one other scientific domain, and competent software engineers with access to large compute resources. People who match this description are few and far between, unfortunately leading to a shrinking pool of possible participants and a loss of experts dedicating their time to solving important problems. Participation is even further restricted in the context of any challenge run on confidential use cases or with sensitive data. Recently, we designed and ran a deep learning challenge to crowd-source the development of an automated labelling system for brain recordings, aiming to advance epilepsy research. A focus of this challenge, run internally in IBM, was the development of a platform that lowers the barrier of entry and therefore mitigates the risk of excluding interested parties from participating. The challenge: enabling wide participation With the goal to run a challenge that mobilises the largest possible pool of participants from IBM (global), we designed a use case around previous work in epileptic seizure prediction [3]. In this “Deep Learning Epilepsy Detection Challenge”, participants were asked to develop an automatic labelling system to reduce the time a clinician would need to diagnose patients with epilepsy. Labelled training and blind validation data for the challenge were generously provided by Temple University Hospital (TUH) [4]. TUH also devised a novel scoring metric for the detection of seizures that was used as basis for algorithm evaluation [5]. In order to provide an experience with a low barrier of entry, we designed a generalisable challenge platform under the following principles: 1. No participant should need to have in-depth knowledge of the specific domain. (i.e. no participant should need to be a neuroscientist or epileptologist.) 2. No participant should need to be an expert data scientist. 3. No participant should need more than basic programming knowledge. (i.e. no participant should need to learn how to process fringe data formats and stream data efficiently.) 4. No participant should need to provide their own computing resources. In addition to the above, our platform should further • guide participants through the entire process from sign-up to model submission, • facilitate collaboration, and • provide instant feedback to the participants through data visualisation and intermediate online leaderboards. The platform The architecture of the platform that was designed and developed is shown in Figure 1. The entire system consists of a number of interacting components. (1) A web portal serves as the entry point to challenge participation, providing challenge information, such as timelines and challenge rules, and scientific background. The portal also facilitated the formation of teams and provided participants with an intermediate leaderboard of submitted results and a final leaderboard at the end of the challenge. (2) IBM Watson Studio [6] is the umbrella term for a number of services offered by IBM. Upon creation of a user account through the web portal, an IBM Watson Studio account was automatically created for each participant that allowed users access to IBM's Data Science Experience (DSX), the analytics engine Watson Machine Learning (WML), and IBM's Cloud Object Storage (COS) [7], all of which will be described in more detail in further sections. (3) The user interface and starter kit were hosted on IBM's Data Science Experience platform (DSX) and formed the main component for designing and testing models during the challenge. DSX allows for real-time collaboration on shared notebooks between team members. A starter kit in the form of a Python notebook, supporting the popular deep learning libraries TensorFLow [8] and PyTorch [9], was provided to all teams to guide them through the challenge process. Upon instantiation, the starter kit loaded necessary python libraries and custom functions for the invisible integration with COS and WML. In dedicated spots in the notebook, participants could write custom pre-processing code, machine learning models, and post-processing algorithms. The starter kit provided instant feedback about participants' custom routines through data visualisations. Using the notebook only, teams were able to run the code on WML, making use of a compute cluster of IBM's resources. The starter kit also enabled submission of the final code to a data storage to which only the challenge team had access. (4) Watson Machine Learning provided access to shared compute resources (GPUs). Code was bundled up automatically in the starter kit and deployed to and run on WML. WML in turn had access to shared storage from which it requested recorded data and to which it stored the participant's code and trained models. (5) IBM's Cloud Object Storage held the data for this challenge. Using the starter kit, participants could investigate their results as well as data samples in order to better design custom algorithms. (6) Utility Functions were loaded into the starter kit at instantiation. This set of functions included code to pre-process data into a more common format, to optimise streaming through the use of the NutsFlow and NutsML libraries [10], and to provide seamless access to the all IBM services used. Not captured in the diagram is the final code evaluation, which was conducted in an automated way as soon as code was submitted though the starter kit, minimising the burden on the challenge organising team. Figure 1: High-level architecture of the challenge platform Measuring success The competitive phase of the "Deep Learning Epilepsy Detection Challenge" ran for 6 months. Twenty-five teams, with a total number of 87 scientists and software engineers from 14 global locations participated. All participants made use of the starter kit we provided and ran algorithms on IBM's infrastructure WML. Seven teams persisted until the end of the challenge and submitted final solutions. The best performing solutions reached seizure detection performances which allow to reduce hundred-fold the time eliptologists need to annotate continuous EEG recordings. Thus, we expect the developed algorithms to aid in the diagnosis of epilepsy by significantly shortening manual labelling time. Detailed results are currently in preparation for publication. Equally important to solving the scientific challenge, however, was to understand whether we managed to encourage participation from non-expert data scientists. Figure 2: Primary occupation as reported by challenge participants Out of the 40 participants for whom we have occupational information, 23 reported Data Science or AI as their main job description, 11 reported being a Software Engineer, and 2 people had expertise in Neuroscience. Figure 2 shows that participants had a variety of specialisations, including some that are in no way related to data science, software engineering, or neuroscience. No participant had deep knowledge and experience in data science, software engineering and neuroscience. Conclusion Given the growing complexity of data science problems and increasing dataset sizes, in order to solve these problems, it is imperative to enable collaboration between people with differences in expertise with a focus on inclusiveness and having a low barrier of entry. We designed, implemented, and tested a challenge platform to address exactly this. Using our platform, we ran a deep-learning challenge for epileptic seizure detection. 87 IBM employees from several business units including but not limited to IBM Research with a variety of skills, including sales and design, participated in this highly technical challenge. 

   more » « less
3. Defining and Assessing Systems Thinking in Diverse Engineering Populations

   Mosyjowski, E.A. ; Daly, S.R. ; Lattuca, L. ( January 2019 , Proceedings of the American Society for Engineering Education Conference)

   Engineers are called to play an important role in addressing the complex problems of our global society, such as climate change and global health care. In order to adequately address these complex problems, engineers must be able to identify and incorporate into their decision making relevant aspects of systems in which their work is contextualized, a skill often referred to as systems thinking. However, within engineering, research on systems thinking tends to emphasize the ability to recognize potentially relevant constituent elements and parts of an engineering problem, rather than how these constituent elements and parts are embedded in broader economic, sociocultural, and temporal contexts and how all of these must inform decision making about problems and solutions. Additionally, some elements of systems thinking, such as an awareness of a particular sociocultural context or the coordination of work among members of a cross-disciplinary team, are not always recognized as core engineering skills, which alienates those whose strengths and passions are related to, for example, engineering systems that consider and impact social change. Studies show that women and minorities, groups underrepresented within engineering, are drawn to engineering in part for its potential to address important social issues. Emphasizing the importance of systems thinking and developing a more comprehensive definition of systems thinking that includes both constituent parts and contextual elements of a system will help students recognize the relevance and value of these other elements of engineering work and support full participation in engineering by a diverse group of students. We provide an overview of our study, in which we are examining systems thinking across a range of expertise to develop a scenario-based assessment tool that educators and researchers can use to evaluate engineering students’ systems thinking competence. Consistent with the aforementioned need to define and study systems thinking in a comprehensive, inclusive manner, we begin with a definition of systems thinking as a holistic approach to problem solving in which linkages and interactions of the immediate work with constituent parts, the larger sociocultural context, and potential impacts over time are identified and incorporated into decision making. In our study, we seek to address two key questions: 1) How do engineers of different levels of education and experience approach problems that require systems thinking? and 2) How do different types of life, educational, and work experiences relate to individuals’ demonstrated level of expertise in solving systems thinking problems? Our study is comprised of three phases. The first two phases include a semi-structured interview with engineering students and professionals about their experiences solving a problem requiring systems thinking and a think-aloud interview in which participants are asked to talk through how they would approach a given engineering scenario and later reflect on the experiences that inform their thinking. Data from these two phases will be used to develop a written assessment tool, which we will test by administering the written instrument to undergraduate and graduate engineering students in our third study phase. Our paper describes our study design and framing and includes preliminary findings from the first phase of our study. 

   more » « less
4. Rodent Parasite Data for the Sevilleta National Wildlife Refuge, New Mexico (1990-1998)

   https://doi.org/10.6073/pasta/b739d2c75e142f5742b7bfe74c97d5c1  

   Duszyinski, Don ( January 2016 , Environmental Data Initiative)

   The 100,000 ha Sevilleta National Wildlife Refuge (SNWR) in central New Mexico lies in a transition zone that straddles several major biomes of the Southwest, including Great Basin Shrub-Steppe, Mogollon Pinon-Juniper Woodland, Great Plains Grassland and Chihuahuan Desert. During 9 years, (1990-1998), collaborating with the University of New Mexico's Long Term Ecological Research (LTER) program, 3,235 rodents (28 species in 4 families) were collected and identified from permanent collecting sites on the 3 major habitat types (grassland, desert/creosote, woodland) on the SNWR. Hosts were necropsied for endoparasites (protozoa [coccidia], helminths) and some ectoparasites. We identified and analyzed all the parasites found in these hosts. By 1998, we had in place the means to easily identify and moniter the parasites from all mammalian hosts caught on the LTER Phase II grant.This is not just another parasite survey; the data we collected was unique for several reasons: 1) This was the first complete inventory of a natural assemblage of parasites from all mammalian (rodent) hosts in 3 different communities, each from a distinctly defined geographic locality (habitat type) over the period of a decade, and beyond; 2) This study was part of a multidisciplinary approach to address conceptual issues of climate change on ecosystem structure and function at multiple scales (individuals, communities, etc) and correlative data from these related studies will strengthen and contribute to the robustness of this data set; 3) As the only parasite study on any of the LTER projects nationwide, it provided an ideal model, and perhaps incentive for parallel longterm studies of parasite communities to be examined in a variety of other habitat types, and from a variety of different perspectives, and other LTER sites in the network.Upon completing the work, we were able to use these long-term data to try to understand the dynamics of natural host-parasite assemblages. Hypotheses were then erected to test/address at least these questions: How do the different parasite communities colonize, mature, climax and senesce over time (or do they?), Do they vary in response to abiotic (climate change) and/or biotic (dispersal, colonization) factors? What temporal and spatial scales, and among what kinds of organisms, do coevolutionary processes influence the community organization of these parasites? Studies of the dynamics of multiple, coexisting species are confined primarily to microtine rodents and have hinted that multiannual cycles tend to be synchronous (Brown and Heske 1990). Are similar patterns seen for the parasites of our desert rodents? Answers to these questions relating to community structure, as well as to questions concerning parasite biodiversity on the SNWR, can be answered paritially or completely by the information we gathered on the parasite species infecting rodents collected on the SNWR. Initial emphasis of our work was on identifying all the parasites collected, by processing 8 consecutive years of parasite data, and on training the undergraduate and graduate students involved in the art of taxonomy and nomenclature of parasitic protozoans and helminths,  to supply some of these answers. 

   more » « less
5. Practical Accuracy Estimation for Efficient Deep Neural Network Testing

   https://doi.org/10.1145/3394112  

   Chen, Junjie ; Wu, Zhuo ; Wang, Zan ; You, Hanmo ; Zhang, Lingming ; Yan, Ming ( October 2020 , ACM Transactions on Software Engineering and Methodology)

   null (Ed.)

   Deep neural network (DNN) has become increasingly popular and DNN testing is very critical to guarantee the correctness of DNN, i.e., the accuracy of DNN in this work. However, DNN testing suffers from a serious efficiency problem, i.e., it is costly to label each test input to know the DNN accuracy for the testing set, since labeling each test input involves multiple persons (even with domain-specific knowledge) in a manual way and the testing set is large-scale. To relieve this problem, we propose a novel and practical approach, called PACE (which is short for P ractical AC curacy E stimation), which selects a small set of test inputs that can precisely estimate the accuracy of the whole testing set. In this way, the labeling costs can be largely reduced by just labeling this small set of selected test inputs. Besides achieving a precise accuracy estimation, to make PACE more practical it is also required that it is interpretable, deterministic, and as efficient as possible. Therefore, PACE first incorporates clustering to interpretably divide test inputs with different testing capabilities (i.e., testing different functionalities of a DNN model) into different groups. Then, PACE utilizes the MMD-critic algorithm, a state-of-the-art example-based explanation algorithm, to select prototypes (i.e., the most representative test inputs) from each group, according to the group sizes, which can reduce the impact of noise due to clustering. Meanwhile, PACE also borrows the idea of adaptive random testing to select test inputs from the minority space (i.e., the test inputs that are not clustered into any group) to achieve great diversity under the required number of test inputs. The two parallel selection processes (i.e., selection from both groups and the minority space) compose the final small set of selected test inputs. We conducted an extensive study to evaluate the performance of PACE based on a comprehensive benchmark (i.e., 24 pairs of DNN models and testing sets) by considering different types of models (i.e., classification and regression models, high-accuracy and low-accuracy models, and CNN and RNN models) and different types of test inputs (i.e., original, mutated, and automatically generated test inputs). The results demonstrate that PACE is able to precisely estimate the accuracy of the whole testing set with only 1.181%∼2.302% deviations, on average, significantly outperforming the state-of-the-art approaches. 

   more » « less