surveyofdnnlearning.pdf
- 1. A Survey of Learning Methods in Deep Neural Networks (DNN) Presented by Ankita Tiwari Hibah Ihsan Muhammad Gaurav Trivedi Department of Electronics and Electrical Engineering IIT Guwahati
- 2. Definition by Tom Mitchell(1998): Machine Learning is the study of algorithms that • improve their performance P • at some task T • with experience E A well-defined learning task is given by <P, T, E>. What is Deep Learning? “Learning is a process by which a system improves its performance from experience.” - Herbert Simon
- 4. Classification of Artificial Intelligence
- 5. Types of Learning Supervised Learning Unsupervised Learning
- 6. When Do We Use Deep Learning DL is used when: • Human expertise does not exist (navigation on mars ) • Human can't explain their expertise (speech recognition) • Models must be customized (personalized medicine) • Model are based on huge amounts of data(genomics) Learning isn't always useful: • There is no need to "learn" to calculate payroll
- 7. 1. What exactly is deep learning ? 2. Why is it generally better than other methods on image, speech and certain other types of data? The short answers • ‘Deep Learning’ means using a neural network with several layers of nodes between input and output, • The combination of layers between input & output perform feature identification and processing in stages, just as our brains seem to.
- 8. Analogy Between Biological and ANN Biological Neuron Artificial Neuron
- 10. 2.7 -8.6 0.002 f(x) 1.4 -2.5 -0.06 x = -0.06×2.7 + 2.5×8.6 + 1.4×0.002 = 21.34 Working of Perceptron
- 11. Deep Learning Architecture • RNN-Recurrent Neural Network, • LSTM-Long Short-Term Memory, • CNN-Convolutional Neural Network, • GRU-Gated Recurrent Units, • RBM-Restricted Boltzmann Machine, • DBN-Deep Belief Network and • GAN-Generative Adversarial Network (GAN) • AE-Auto-Encoder
- 12. Proposed Study • Per Bayes approach, we started with prior & available information collected from Centers for Disease Control and Prevention (https://covid.cdc.gov/covid-data- tracker/#cases_casesper100klast7days), • John Hopkins university GitHub page (https://github.com/CSSEGISandData/COVID- 19/tree/master/csse_covid_19_data), and we extracted the COVID- 19 raw data from the above Johns Hopkins university Github repository using the following data extraction steps for data analysis.
- 13. Step 1: We downloaded the COVID-19 .csv raw dataset from the above JHU GitHub page Step 2: Raw (.csv) dataset loaded into ‘staging tables’ and extract the common date list (for example only current day information is extracted) Step 3: We merged the following ‘raw data confirmed cases’, ‘raw data confirmed deaths’, and ‘raw data confirmed recovered’ into ‘target table’ Step 4: We created dataset/data-frame using the ‘target table’ data for data analysis Step 5: We aggregated data into region wise and group them by date and region. After that, we added day-wise new cases, new deaths and new recovered by deducting the corresponding cumulative data on the previous day. And we updated the incidence rate using the posterior distribution for an estimate the incidence rate of coronavirus disease using the below statistical model
- 15. Conclusion • Humanity is looking forward to the prospects of AI to resolve the challenges that are seen insurmountable to date. • To illustrate, in health care, DL is being increasingly tested for the early diagnosis of disorders and diseases, including Alzheimer's, Parkinson's, developmental disorders, etc. • Deep learning is growing exponentially demonstrating its success and versatility of applications in diverse areas. • In addition, the rapidly improved accuracy rates clearly exhibit the relevance and prospects for deep learning advancement. In the evolution of DL, the hierarchy of layers, learning models and algorithms are critical key factors to evolve an efficacious implementation with deep Learning.
- 16. References 1. Arora, Gunjan, Jayadev Joshi, Rahul Shubhra Mandal, Nitisha Shrivastava and Richa Virmani, 18 August 2021. Artificial Intelligence in Surveillance, Diagnosis, Drug Discovery and Vaccine Development against COVID-19. Pathogens, 2021, 10, 1048. 2. Schmidhuber, J. Deep Learning in Neural Networks: An Overview. Neural Netw. 2015, 61, 85–117. 3. Shaveta Dargan: Munish Kumar and Maruthi Rohit Ayyagari, A survey of Deep Learning and its applications: A New Paradigm to Machine Learning: Research Gate. July 2019 4. Deng J., Dong W., Socher R., Li L.-J., Li K. and Fei-Fei L. Imagenet: A large-scale hierarchical image database. 2009 IEEE Conference on Computer Vision and Pattern Recognition, IEEE (2009), pp. 248-255 5. Chen, Y.; Luo, T.; Liu, S.; Zhang, S.; He, L.; Wang, J.; Li, L.; Chen, T.; Xu, Z.; Sun, N.; et al. Dadiannao: A machine-learning supercomputer. In Proceedings of the 47th Annual IEEE/ACM International Symposium on Microarchitecture, Cambridge, UK, 13–17 December 2014; pp. 609–622. 6. Coates, A., Huval, B., Wang, T., Wu, D.J., Catanzaro, B. and Ng, A.Y., 2013. Deep learning with COTS HPC systems. In: ICML. Google Scholar 7. Abadi, M., Barham, P., Chen, J., Chen, Z., Davis, A., Dean, J., Devin, M., Ghemawat, S., Irving, G., Isard, M., Kudlur, M., Levenberg, J., Monga, R., Moore, S., Murray, D.G., Steiner, B., Tucker, P.A., Vasudevan, V., Warden, P., Wicke, M., Yu, Y. and Zhang, X., 2016. 8. TensorFlow: A system for large-scale machine learning. In: OSDI.Alzubaidi Laith, Jinglan Zhang , Amjad J. Humaidi , Ayad AlDujaili , Ye Duan , Omran AlShamma , J. Santamaría , Mohammed A. Fadhel , Muthana AlAmidie and Laith Farhan /Alzubaidi et al. Review of deep learning: concepts, CNN architectures, challenges, applications, future directions. J Big Data (2021) 8:53 9. Goodfellow, I.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A. and Bengio, Y. Generative adversarial nets. In Advances in Neural Information Processing Systems; The MIT Press:Cambridge, MA, USA, 2014; pp. 2672–2680 10. Vondrick, C.; Pirsiavash, H. and Torralba, A. Generating videos with scene dynamics. In Advances in Neural Information Processing Systems; MIT Press: Cambridge, MA, USA, 2016; pp. 613–621. 11. Zahangir Alom Md; Tarek M. Taha; Chris Yakopcic; Stefan Westberg ; Paheding Sidike; Mst Shamima Nasrin; Mahmudul Hasan; Brian C. Van Essen; Abdul A. S. Awwal and Vijayan K. Asari: A State-of-the-Art Survey on Deep Learning Theory and Architectures. Electronics MDPI. Electronics 2019, 8, 292; doi:10.3390/electronics8030292 12. Kendall, A. and Gal, Y. What uncertainties do we need in bayesian deep learning for computer vision? In Advances in Neural Information Processing Systems
- 17. Thank You!