• DL Session #6: Character Recognition by Deep Learning: An Enterprise solution

    Deep Learning is a new course organized by the Austin ACM SIGKDD Chapter and sponsored by Visa.

    Session #6: Character Recognition by Deep Learning: An Enterprise solution

    Abstract:
    Ease-of-use analytics at scale is the holy grail of industrial strength machine learning. While there have been advances in APIs, algorithms, and user interfaces, building an end to end work flow involving data ingestion, data preparation, model training, model scoring, and visualization received limited investment and effort producing only marginal innovation. This talk outlines a proof of concept that demonstrates an analytical workflow that integrates multiple analytical tools and techniques for image recognition. The solution combines Relational Databases and Machine Learning (teradata), Deep Learning (TensorFlow), Distributed File System (HDFS), Graphical Processing Units, and user interface tools over a communication fabric (teradata QueryGrid). In particular, we demonstrate hand written word recognition through an application of Convolutional Neural Networks in TensorFlow and teradata's custom analytical functions to recognize first names and last names in the payee field section of financial negotiable instruments such as in a cheque. We hope that this talk will serve as a guide to successful implementation of analytical workflows in production.

    Speaker: Thiagarajan Ramakrishnan
    Thiagu is a Senior Software Engineer at Teradata. His love for technology began right from his first PC, and endless possibilities it provided. He found his passion for Data Science and Engineering field during his Undergraduate Program. Later, he went on to get his MS in Computer Science field, specializing primarily on Data Science and Intelligent systems track. Thiagu loves building products which can solve real time problems to the users across the globe. In addition to that, he's also passionate about problem solving and had taught programming courses as part of his CS Outreach program during his grad school day

    Agenda:

    6:30 Pizza + Networking
    7:00 Presentation + QA

    Location:
    Visa (12301 Research Blvd, Bldg 3, Austin, TX 78759) will sponsor the classes and provide pizza and drinks.

    Parking is available in front of the Visa building. If you cannot find a spot, you can park in the garage behind the building.

    RSVP:

    • Seating is limited to the first 150 to RSVP each week.

    • Please bring a picture ID and arrive early to assist with the sign-in process. Please arrive no later than 6:45PM.

    • Please send your full name when RSVP. Without a full name, we cannot prepare a badge for you. This will cause a significant delay.

    • RSVP will end on 10/15 at 1PM.

    • Please let me know if you have any questions about RSVP.

    3
  • Approachable Data Science: An Introduction

    Galvanize - Austin

    To complement our more advanced topics, we will be starting an introductory course to help those in our group who want to get started in building machine learning models. This will be a six part course covered in 1-1.5 hour workshops. Each session will take the form of part instructor led training and part hands-on exercises. We will cover how you start and organize a project, exploring the data, preparing the data, building the model, evaluating the model and publishing the model so others can use your work. We will plan to have a session every 2 weeks. At the end of this course if you attend at least 75% of the sessions then you will receive a certificate for the course from the ACM in the final session. I will announce the location once I get a sense of the size of the group. This course is at no cost to you and you do not need to be a member to attend.

    Course Description:
    Interested in Data Science but don't have a computer science or math background? We are taking the intimidation out of this subject. This workshop will help you understand the building blocks of a data science project, introduce steps to take before getting started, and will familiarize you with tools to create classifiers and predictors.
    Approachable Data Science will give you simple approaches to going beyond tools like Microsoft Excel by walking you through a typical project involving data science. This will include strategies around structuring a project for success, using tools to simplify implementation, and how to clearly share results with your team.
    For example, one of the models we will build addresses a problem that many businesses face- Churn. We will use billing data to train a model to see if we can predict customers that will terminate. This real world example provides space to explore the data involved and build a machine learning model with a data science toolkit. Then you will evaluate your results and learn how to share them with others.

    Suggested Experience Level:
    You do not need a programming background for this class. You should however be comfortable with basic data manipulation in spreadsheets and you should be familiar with basic descriptive statistics concepts. For example, you should know how to get the sum, mean, minimum and maximum from a set of numbers. We will be using machine learning workbench called KNIME which is free to download and will not require code to use.

    What to bring:
    You should bring your own laptop (either Windows/Mac) so you can use the software to run the exercises we provide. Show up prepared to build a real machine learning model!

    Developed by Blacklight Solutions:
    Blacklight Solutions works with businesses to unlock the power of the data they own, generate, or collect in their business. Blacklight Solutions enables businesses to create products with their owned data that expands customer engagement and increases revenue. In turn we help our clients put cutting edge experiences in front of their customers that generate insights and increase transparency.

    The Instructor:
    Chance Coble has been helping people get value from their data for over 20 years. He has been recruited to lead predictive analytics solutions at some of the most successful organizations in the world. He has been selected by the United States government and other governments around the world to implement cutting edge solutions in machine learning and artificial intelligence. Mr. Coble is currently President and CEO of Blacklight Solutions where he spends his days finding practical approaches for small and mid-sized organizations to execute on and benefit from their analytics and AI strategies. He holds a B.S. in Computer Science from the University of Texas at Austin and an MS in Biomedical Informatics from the University of Texas at Houston.

    8
  • Deep Learning Course. Session #5: Tensors, Pytorch & fastai

    Deep Learning is a new course organized by the Austin ACM SIGKDD Chapter and sponsored by Visa.

    Session #5: Tensors, Pytorch & fastai.

    Abstract:
    Neural networks are a foundational driver in deep learning initiatives. In this talk, I will explore critical developments in neural networks that have enabled them to become such a key technology. I will cover each development both from the perspective of the the key technical concepts that were behind the innovation, the hurdles it helped overcome, and how it fits into the development of neural network architectures. This talk also will include an overview of some of the structures and principles behind the current deep learning architectures, for example CNNs, RNNs, embeddings, sequence models.

    Speaker: Dr. Misty Nodine
    Misty is currently an independent consultant in all things data, especially Data Science and Data Architecture. She is also in the very early stages of founding a non-profit.

    Misty enjoys supporting startups. As Lead Data Scientist at StepOne, Inc., she was responsible for the design and architecture of the data science and machine learning portions of StepOne’s products. At Elastic Knowledge, Misty consulted with early startup companies on how they could leverage machine learning technologies to improve their products. She provided focused product design support for selected ML features. She ran Machine Learning Monday Lunch at Tech Ranch to help educate entrepreneurs on machine learning.

    Misty’s background in government research includes positions at Telcordia Applied Research Labs, MCC, and Bolt, Beranek and Newman. Her main research thrusts included subjects as varied as agent-based systems, knowledge management and decision support, collaborative teamwork support and network protocols. As an example, at Telcordia she worked on a system to automatically analyze problems in networks and in agent-based systems (faults, security breaches, performance issues), and suggest/automate corrective responses. Misty's work at MCC included being the technical lead for the InfoSleuth team that did the core architecture for query processing over information sources distributed across the Internet.

    Misty received her Ph.D. in Computer Science from Brown University. She received her S.B. and S.M. in EECS from Massachusetts Institute of Technology. She is an author of over 30 peer-reviewed technical papers.

    Agenda:

    6:30 Pizza + Networking
    7:00 Presentation + QA

    Certificate: An official ACM certificate will be awarded to those who complete the course. In order to qualify, you need to attend at least 65% of the sessions.

    Location:
    Visa (12301 Research Blvd, Bldg 3, Austin, TX 78759) will sponsor the classes and provide pizza and drinks.

    Parking is available in front of the Visa building. If you cannot find a spot, you can park in the garage behind the building.

    RSVP:

    • Seating is limited to the first 150 to RSVP each week.

    • Please bring a picture ID and arrive early to assist with the sign-in process. Please arrive no later than 6:45PM.

    • Please send your full name when RSVP. Without a full name, we cannot prepare a badge for you. This will cause a significant delay.

    • RSVP will end on 09/24 at 1PM.

    • Please let me know if you have any questions about RSVP.

    2
  • Approachable Data Science: An Introduction

    Galvanize - Austin

    This course is now a joint event between the ACM and Women in Data Science (https://www.meetup.com/Women-in-Data-Science-ATX/).

    To complement our more advanced topics, we will be starting an introductory course to help those in our group who want to get started in building machine learning models. This will be a six part course covered in 1-1.5 hour workshops. Each session will take the form of part instructor led training and part hands-on exercises. We will cover how you start and organize a project, exploring the data, preparing the data, building the model, evaluating the model and publishing the model so others can use your work. We will plan to have a session every 2 weeks. At the end of this course if you attend at least 75% of the sessions then you will receive a certificate for the course from the ACM in the final session. I will announce the location once I get a sense of the size of the group. This course is at no cost to you and you do not need to be a member to attend.

    Course Description:
    Interested in Data Science but don't have a computer science or math background? We are taking the intimidation out of this subject. This workshop will help you understand the building blocks of a data science project, introduce steps to take before getting started, and will familiarize you with tools to create classifiers and predictors.
    Approachable Data Science will give you simple approaches to going beyond tools like Microsoft Excel by walking you through a typical project involving data science. This will include strategies around structuring a project for success, using tools to simplify implementation, and how to clearly share results with your team.
    For example, one of the models we will build addresses a problem that many businesses face- Churn. We will use billing data to train a model to see if we can predict customers that will terminate. This real world example provides space to explore the data involved and build a machine learning model with a data science toolkit. Then you will evaluate your results and learn how to share them with others.

    Suggested Experience Level:
    You do not need a programming background for this class. You should however be comfortable with basic data manipulation in spreadsheets and you should be familiar with basic descriptive statistics concepts. For example, you should know how to get the sum, mean, minimum and maximum from a set of numbers. We will be using machine learning workbench called KNIME which is free to download and will not require code to use.

    What to bring:
    You should bring your own laptop (either Windows/Mac) so you can use the software to run the exercises we provide. Show up prepared to build a real machine learning model!

    Developed by Blacklight Solutions:
    Blacklight Solutions works with businesses to unlock the power of the data they own, generate, or collect in their business. Blacklight Solutions enables businesses to create products with their owned data that expands customer engagement and increases revenue. In turn we help our clients put cutting edge experiences in front of their customers that generate insights and increase transparency.

    The Instructor:
    Chance Coble has been helping people get value from their data for over 20 years. He has been recruited to lead predictive analytics solutions at some of the most successful organizations in the world. He has been selected by the United States government and other governments around the world to implement cutting edge solutions in machine learning and artificial intelligence. Mr. Coble is currently President and CEO of Blacklight Solutions where he spends his days finding practical approaches for small and mid-sized organizations to execute on and benefit from their analytics and AI strategies. He holds a B.S. in Computer Science from the University of Texas at Austin and an MS in Biomedical Informatics from the University of Texas at Houston.

    2
  • Approachable Data Science: An Introduction

    Galvanize - Austin

    This course is now a joint event between the ACM and Women in Data Science (https://www.meetup.com/Women-in-Data-Science-ATX/).

    To complement our more advanced topics, we will be starting an introductory course to help those in our group who want to get started in building machine learning models. This will be a six part course covered in 1-1.5 hour workshops. Each session will take the form of part instructor led training and part hands-on exercises. We will cover how you start and organize a project, exploring the data, preparing the data, building the model, evaluating the model and publishing the model so others can use your work. We will plan to have a session every 2 weeks. At the end of this course if you attend at least 75% of the sessions then you will receive a certificate for the course from the ACM in the final session. I will announce the location once I get a sense of the size of the group. This course is at no cost to you and you do not need to be a member to attend.

    Course Description:
    Interested in Data Science but don't have a computer science or math background? We are taking the intimidation out of this subject. This workshop will help you understand the building blocks of a data science project, introduce steps to take before getting started, and will familiarize you with tools to create classifiers and predictors.
    Approachable Data Science will give you simple approaches to going beyond tools like Microsoft Excel by walking you through a typical project involving data science. This will include strategies around structuring a project for success, using tools to simplify implementation, and how to clearly share results with your team.
    For example, one of the models we will build addresses a problem that many businesses face- Churn. We will use billing data to train a model to see if we can predict customers that will terminate. This real world example provides space to explore the data involved and build a machine learning model with a data science toolkit. Then you will evaluate your results and learn how to share them with others.

    Suggested Experience Level:
    You do not need a programming background for this class. You should however be comfortable with basic data manipulation in spreadsheets and you should be familiar with basic descriptive statistics concepts. For example, you should know how to get the sum, mean, minimum and maximum from a set of numbers. We will be using machine learning workbench called KNIME which is free to download and will not require code to use.

    What to bring:
    You should bring your own laptop (either Windows/Mac) so you can use the software to run the exercises we provide. Show up prepared to build a real machine learning model!

    Developed by Blacklight Solutions:
    Blacklight Solutions works with businesses to unlock the power of the data they own, generate, or collect in their business. Blacklight Solutions enables businesses to create products with their owned data that expands customer engagement and increases revenue. In turn we help our clients put cutting edge experiences in front of their customers that generate insights and increase transparency.

    The Instructor:
    Chance Coble has been helping people get value from their data for over 20 years. He has been recruited to lead predictive analytics solutions at some of the most successful organizations in the world. He has been selected by the United States government and other governments around the world to implement cutting edge solutions in machine learning and artificial intelligence. Mr. Coble is currently President and CEO of Blacklight Solutions where he spends his days finding practical approaches for small and mid-sized organizations to execute on and benefit from their analytics and AI strategies. He holds a B.S. in Computer Science from the University of Texas at Austin and an MS in Biomedical Informatics from the University of Texas at Houston.

    1
  • Approachable Data Science: An Introduction

    Galvanize - Austin

    This course is now a joint event between the ACM and Women in Data Science (https://www.meetup.com/Women-in-Data-Science-ATX/).

    To complement our more advanced topics, we will be starting an introductory course to help those in our group who want to get started in building machine learning models. This will be a six part course covered in 1-1.5 hour workshops. Each session will take the form of part instructor led training and part hands-on exercises. We will cover how you start and organize a project, exploring the data, preparing the data, building the model, evaluating the model and publishing the model so others can use your work. We will plan to have a session every 2 weeks. At the end of this course if you attend at least 75% of the sessions then you will receive a certificate for the course from the ACM in the final session. I will announce the location once I get a sense of the size of the group. This course is at no cost to you and you do not need to be a member to attend.

    Course Description:
    Interested in Data Science but don't have a computer science or math background? We are taking the intimidation out of this subject. This workshop will help you understand the building blocks of a data science project, introduce steps to take before getting started, and will familiarize you with tools to create classifiers and predictors.
    Approachable Data Science will give you simple approaches to going beyond tools like Microsoft Excel by walking you through a typical project involving data science. This will include strategies around structuring a project for success, using tools to simplify implementation, and how to clearly share results with your team.
    For example, one of the models we will build addresses a problem that many businesses face- Churn. We will use billing data to train a model to see if we can predict customers that will terminate. This real world example provides space to explore the data involved and build a machine learning model with a data science toolkit. Then you will evaluate your results and learn how to share them with others.

    Suggested Experience Level:
    You do not need a programming background for this class. You should however be comfortable with basic data manipulation in spreadsheets and you should be familiar with basic descriptive statistics concepts. For example, you should know how to get the sum, mean, minimum and maximum from a set of numbers. We will be using machine learning workbench called KNIME which is free to download and will not require code to use.

    What to bring:
    You should bring your own laptop (either Windows/Mac) so you can use the software to run the exercises we provide. Show up prepared to build a real machine learning model!

    Developed by Blacklight Solutions:
    Blacklight Solutions works with businesses to unlock the power of the data they own, generate, or collect in their business. Blacklight Solutions enables businesses to create products with their owned data that expands customer engagement and increases revenue. In turn we help our clients put cutting edge experiences in front of their customers that generate insights and increase transparency.

    The Instructor:
    Chance Coble has been helping people get value from their data for over 20 years. He has been recruited to lead predictive analytics solutions at some of the most successful organizations in the world. He has been selected by the United States government and other governments around the world to implement cutting edge solutions in machine learning and artificial intelligence. Mr. Coble is currently President and CEO of Blacklight Solutions where he spends his days finding practical approaches for small and mid-sized organizations to execute on and benefit from their analytics and AI strategies. He holds a B.S. in Computer Science from the University of Texas at Austin and an MS in Biomedical Informatics from the University of Texas at Houston.

    3
  • Approachable Data Science: An Introduction

    Galvanize - Austin

    This course is now a joint event between the ACM and Women in Data Science (https://www.meetup.com/Women-in-Data-Science-ATX/).

    To complement our more advanced topics, we will be starting an introductory course to help those in our group who want to get started in building machine learning models. This will be a six part course covered in 1-1.5 hour workshops. Each session will take the form of part instructor led training and part hands-on exercises. We will cover how you start and organize a project, exploring the data, preparing the data, building the model, evaluating the model and publishing the model so others can use your work. We will plan to have a session every 2 weeks. At the end of this course if you attend at least 75% of the sessions then you will receive a certificate for the course from the ACM in the final session. I will announce the location once I get a sense of the size of the group. This course is at no cost to you and you do not need to be a member to attend.

    Course Description:
    Interested in Data Science but don't have a computer science or math background? We are taking the intimidation out of this subject. This workshop will help you understand the building blocks of a data science project, introduce steps to take before getting started, and will familiarize you with tools to create classifiers and predictors.
    Approachable Data Science will give you simple approaches to going beyond tools like Microsoft Excel by walking you through a typical project involving data science. This will include strategies around structuring a project for success, using tools to simplify implementation, and how to clearly share results with your team.
    For example, one of the models we will build addresses a problem that many businesses face- Churn. We will use billing data to train a model to see if we can predict customers that will terminate. This real world example provides space to explore the data involved and build a machine learning model with a data science toolkit. Then you will evaluate your results and learn how to share them with others.

    Suggested Experience Level:
    You do not need a programming background for this class. You should however be comfortable with basic data manipulation in spreadsheets and you should be familiar with basic descriptive statistics concepts. For example, you should know how to get the sum, mean, minimum and maximum from a set of numbers. We will be using machine learning workbench called KNIME which is free to download and will not require code to use.

    What to bring:
    You should bring your own laptop (either Windows/Mac) so you can use the software to run the exercises we provide. Show up prepared to build a real machine learning model!

    Developed by Blacklight Solutions:
    Blacklight Solutions works with businesses to unlock the power of the data they own, generate, or collect in their business. Blacklight Solutions enables businesses to create products with their owned data that expands customer engagement and increases revenue. In turn we help our clients put cutting edge experiences in front of their customers that generate insights and increase transparency.

    The Instructor:
    Chance Coble has been helping people get value from their data for over 20 years. He has been recruited to lead predictive analytics solutions at some of the most successful organizations in the world. He has been selected by the United States government and other governments around the world to implement cutting edge solutions in machine learning and artificial intelligence. Mr. Coble is currently President and CEO of Blacklight Solutions where he spends his days finding practical approaches for small and mid-sized organizations to execute on and benefit from their analytics and AI strategies. He holds a B.S. in Computer Science from the University of Texas at Austin and an MS in Biomedical Informatics from the University of Texas at Houston.

    2
  • DL Course Session #4: Leveraging LSTMs for Natural Language Generation

    Deep Learning is a new course organized by the Austin ACM SIGKDD Chapter and sponsored by Visa.

    This talk will be given by Dr. Jennifer Davis . The focus would be LSTM, BiLSTM and their use in natural language generation

    Speaker Bio:
    Dr. Jennifer Davis is a Ph.D. level, licensed data scientist and applied AI scientist who founded Data Bot Box in 2017. She has a wide variety of experience including algorithm development in Python and MatLab, testing algorithmic ensembles in MatLab, and converting to Python or R. Dr. Davis, is a member of the Association for Computational Linguistics, and the Association for Computing Machinery. Her track record includes, provisional patents, and publications. Some of her clients have successfully obtained investor funds, admission to incubators, or won innovation awards. Dr. Davis has worked as a senior managing consultant at IBM Watson, and as a R&D Principal at Chaotic Moon Research Lab, Accenture Interactive. Her work has been featured at SXSW, IEEE, and in peer-reviewed publications. Jennifer likes to push the boundaries of what is possible, and is currently exploring natural language generation, computer vision, and IoT projects with various clients. Data Bot Box accepts select clients for research project containing business-driven objectives. Feel free to approach with questions after the talk.

    Agenda:
    6:30 Pizza + Networking
    7:00 Presentation + QA

    Certificate: An official ACM certificate will be awarded to those who complete the course. In order to qualify, you need to attend at least 65% of the sessions.

    Location:
    Visa (12301 Research Blvd, Bldg 3, Austin, TX 78759) will sponsor the classes and provide pizza and drinks.

    Parking is available in front of the Visa building. If you cannot find a spot, you can park in the garage behind the building.

    RSVP:

    • Seating is limited to the first 150 to RSVP each week.

    • Please bring a picture ID and arrive early to assist with the sign-in process. Please arrive no later than 6:45PM.

    • Please send your full name when RSVP. Without a full name, we cannot prepare a badge for you. This will cause a significant delay.

    • RSVP will end on 08/27 at 1PM.

    • Please let me know if you have any questions about RSVP.

    2
  • Approachable Data Science: An Introduction

    Galvanize - Austin

    This course is now a joint event between the ACM and Women in Data Science (https://www.meetup.com/Women-in-Data-Science-ATX/).

    To complement our more advanced topics, we will be starting an introductory course to help those in our group who want to get started in building machine learning models. This will be a six part course covered in 1-1.5 hour workshops. Each session will take the form of part instructor led training and part hands-on exercises. We will cover how you start and organize a project, exploring the data, preparing the data, building the model, evaluating the model and publishing the model so others can use your work. We will plan to have a session every 2 weeks. At the end of this course if you attend at least 75% of the sessions then you will receive a certificate for the course from the ACM in the final session. I will announce the location once I get a sense of the size of the group. This course is at no cost to you and you do not need to be a member to attend.

    Course Description:
    Interested in Data Science but don't have a computer science or math background? We are taking the intimidation out of this subject. This workshop will help you understand the building blocks of a data science project, introduce steps to take before getting started, and will familiarize you with tools to create classifiers and predictors.
    Approachable Data Science will give you simple approaches to going beyond tools like Microsoft Excel by walking you through a typical project involving data science. This will include strategies around structuring a project for success, using tools to simplify implementation, and how to clearly share results with your team.
    For example, one of the models we will build addresses a problem that many businesses face- Churn. We will use billing data to train a model to see if we can predict customers that will terminate. This real world example provides space to explore the data involved and build a machine learning model with a data science toolkit. Then you will evaluate your results and learn how to share them with others.

    Suggested Experience Level:
    You do not need a programming background for this class. You should however be comfortable with basic data manipulation in spreadsheets and you should be familiar with basic descriptive statistics concepts. For example, you should know how to get the sum, mean, minimum and maximum from a set of numbers. We will be using machine learning workbench called KNIME which is free to download and will not require code to use.

    What to bring:
    You should bring your own laptop (either Windows/Mac) so you can use the software to run the exercises we provide. Show up prepared to build a real machine learning model!

    Developed by Blacklight Solutions:
    Blacklight Solutions works with businesses to unlock the power of the data they own, generate, or collect in their business. Blacklight Solutions enables businesses to create products with their owned data that expands customer engagement and increases revenue. In turn we help our clients put cutting edge experiences in front of their customers that generate insights and increase transparency.

    The Instructor:
    Chance Coble has been helping people get value from their data for over 20 years. He has been recruited to lead predictive analytics solutions at some of the most successful organizations in the world. He has been selected by the United States government and other governments around the world to implement cutting edge solutions in machine learning and artificial intelligence. Mr. Coble is currently President and CEO of Blacklight Solutions where he spends his days finding practical approaches for small and mid-sized organizations to execute on and benefit from their analytics and AI strategies. He holds a B.S. in Computer Science from the University of Texas at Austin and an MS in Biomedical Informatics from the University of Texas at Houston.

    11