diff --git a/2019-01-29_Tibbles_Oh_My/README.md b/2019-01-29_Tibbles_Oh_My/README.md index 2c054e7..aabc3d7 100644 --- a/2019-01-29_Tibbles_Oh_My/README.md +++ b/2019-01-29_Tibbles_Oh_My/README.md @@ -1,6 +1,7 @@ -#Speaker: Ryan Benz +# Tibbles, and list columns, and nested data frames, oh my! -##Title: Tibbles, and list columns, and nested data frames, oh my! +* 2019-01-29 +* Speaker: Ryan Benz -###Abstract -Tibbles are a "modern reimagining of the data.frame" and are a big part of the tidyverse, but what are they really, how are they different from standard data frames, any why should you care about using them? In this talk, I'll dive into tibbles and show how they can be used in several interesting ways taking advantage of list columns and nested data frames. Regardless of what you use R for, learning to harness the power of the tibble is an essential part of doing effective data analysis in the tidyverse. \ No newline at end of file +## Abstract +Tibbles are a "modern reimagining of the data.frame" and are a big part of the tidyverse, but what are they really, how are they different from standard data frames, any why should you care about using them? In this talk, I'll dive into tibbles and show how they can be used in several interesting ways taking advantage of list columns and nested data frames. Regardless of what you use R for, learning to harness the power of the tibble is an essential part of doing effective data analysis in the tidyverse. diff --git a/2019-01-29_working_with_tidymodels/README.md b/2019-01-29_working_with_tidymodels/README.md index 047f751..ba1402f 100644 --- a/2019-01-29_working_with_tidymodels/README.md +++ b/2019-01-29_working_with_tidymodels/README.md @@ -1,6 +1,8 @@ -#Speaker: Emil Hansen +# Working with tidymodels -##Title: Working with tidymodels +* 2019-01-29 +* Speaker: Emil Hansen -###Abstract + +## Abstract Tidymodels is a “meta-package” in the same way as tidyverse, but with a focus on modeling and statistical analysis. This talk will go through how to use tidymodels to do modeling in a tidy fashion. diff --git a/2019-02-26_opinion_mining/README.md b/2019-02-26_opinion_mining/README.md index 26a65ca..ebf9eb0 100644 --- a/2019-02-26_opinion_mining/README.md +++ b/2019-02-26_opinion_mining/README.md @@ -1,6 +1,7 @@ -#Speaker: Yemi Odeyemi +# Opinion Mining of Climate Change Research -##Title: Opinion Mining of Climate Change Research +* 2019-02-26 +* Speaker: Yemi Odeyemi -###Abstract -Climate is defined as the long-term impact of weather on a specific locality. Statistically, climate is defined as the average weather for a specific area over a defined period, usually over three to four decades. Over the years we have seen the socioeconomic impacts of climate change on public health, agriculture/nutrition, nutrient cycle, migration, local economy. In recent times, there has been a huge debate on the polarity of climate change impact, with energy companies arguing against the impact of climate change while the government regulators arguing for. In this study we assessed the sentiments/opinions of general discourse of NSF funded climate change research to ascertain if funding source influences the outcome of the research. \ No newline at end of file +## Abstract +Climate is defined as the long-term impact of weather on a specific locality. Statistically, climate is defined as the average weather for a specific area over a defined period, usually over three to four decades. Over the years we have seen the socioeconomic impacts of climate change on public health, agriculture/nutrition, nutrient cycle, migration, local economy. In recent times, there has been a huge debate on the polarity of climate change impact, with energy companies arguing against the impact of climate change while the government regulators arguing for. In this study we assessed the sentiments/opinions of general discourse of NSF funded climate change research to ascertain if funding source influences the outcome of the research. diff --git a/2019-02-26_xaringan_tutorial/README.md b/2019-02-26_xaringan_tutorial/README.md index 8f72d37..27b4d95 100755 --- a/2019-02-26_xaringan_tutorial/README.md +++ b/2019-02-26_xaringan_tutorial/README.md @@ -1,5 +1,8 @@ # XaringanTutorial +* 2019-02-26 +* Speaker: Zhi Yang + A quick tutorial on making slides using xaringan diff --git a/2019-03-26_Deep_Learning_Using_R_and_TensorFlow/README.md b/2019-03-26_Deep_Learning_Using_R_and_TensorFlow/README.md index 0b0eb3b..77ffa9b 100644 --- a/2019-03-26_Deep_Learning_Using_R_and_TensorFlow/README.md +++ b/2019-03-26_Deep_Learning_Using_R_and_TensorFlow/README.md @@ -1,10 +1,11 @@ -#Speaker: Ash Pahwa +# Deep Learning using R and TensorFlow (Part 1) -##Title: Deep Learning using R and TensorFlow (Part 1) +* 2019-03-26 +* Speaker: Ash Pahwa -###Abstract: +## Abstract: Google released TensorFlow software in 2015 targeted for the Deep Learning applications. Two primary misconceptions about TensorFlow are as follows. First, it can only be used with Python and second, it can only be used for building deep learning applications This talk will demystify these concepts. First, TensorFlow can also be used with R and any programming language. Second, TensorFlow’s main objective is to solve the mathematical problem (including differential equations) which contains multi-dimensional arrays (or tensors). Deep Learning Neural Networks is just another application of TensorFlow. -In this first part, the fundamental architecture of TensorFlow will be covered. TensorFlow works on the concept of DAG (Directed Acyclic Graph). First, a DAG of the mathematical problem is built, and then the DAG is executed in the most efficient way using CPU, GPU or TPUs. The interface between R and TensorFlow will be discussed by building Neural Networks using R and TensorFlow. \ No newline at end of file +In this first part, the fundamental architecture of TensorFlow will be covered. TensorFlow works on the concept of DAG (Directed Acyclic Graph). First, a DAG of the mathematical problem is built, and then the DAG is executed in the most efficient way using CPU, GPU or TPUs. The interface between R and TensorFlow will be discussed by building Neural Networks using R and TensorFlow. diff --git a/2019-03-26_unit_testing/README.md b/2019-03-26_unit_testing/README.md index deffcef..168d28d 100644 --- a/2019-03-26_unit_testing/README.md +++ b/2019-03-26_unit_testing/README.md @@ -1,8 +1,10 @@ -#Speaker: John Peach +# It’s not that we don’t test our code, it’s that we don’t store our tests -##Title: It’s not that we don’t test our code, it’s that we don’t store our tests +* 2019-03-26 +* Speaker: John Peach -###Abstract: + +## Abstract: Software testing is important and we all do it. However, it is often not done in a principled way. It is standard practice in the software development community but is less common in analytics. testthat is a testing framework for R that is similar to other unit testing systems that you may know. It seamlessly integrates into the package development workflow so that you can test early and often. testthat decreases code development and maintenance frustration. Unit testing will help you develop better code structure by making your code easier to test. Thus, lazy development becomes better development. It allows you to quickly debug issues by isolating the source of a bug faster so that you can get back to the ‘real’ coding. It also allows you to have confidence that your changes did not break existing code. diff --git a/2019-04-29_Deep_Learning_Using_R_and_TensorFlow/README.md b/2019-04-29_Deep_Learning_Using_R_and_TensorFlow/README.md index b548378..8da3fb4 100644 --- a/2019-04-29_Deep_Learning_Using_R_and_TensorFlow/README.md +++ b/2019-04-29_Deep_Learning_Using_R_and_TensorFlow/README.md @@ -1,10 +1,11 @@ -#Speaker: Ash Pahwa +# Deep Learning using R and TensorFlow (Part 2) -##Title: Deep Learning using R and TensorFlow (Part 2) +* 2019-04-29 +* Speaker: Ash Pahwa -###Abstract: +## Abstract: Google released TensorFlow software in 2015 targeted for the Deep Learning applications. Two primary misconceptions about TensorFlow are as follows. First, it can only be used with Python and second, it can only be used for building deep learning applications This talk will demystify these concepts. First, TensorFlow can also be used with R and any programming language. Second, TensorFlow’s main objective is to solve the mathematical problem (including differential equations) which contains multi-dimensional arrays (or tensors). Deep Learning Neural Networks is just another application of TensorFlow. -In this second part, neural networks optimization techniques will be discussed. The implementation of Gradient Descent and Adam (Adaptive Momentum Estimation) optimization techniques will be covered using R and TensorFlow. \ No newline at end of file +In this second part, neural networks optimization techniques will be discussed. The implementation of Gradient Descent and Adam (Adaptive Momentum Estimation) optimization techniques will be covered using R and TensorFlow. diff --git a/2019-05-28_OCRUG_Hackathon_Captain_Rrrr/README.md b/2019-05-28_OCRUG_Hackathon_Captain_Rrrr/README.md index e4ca0f8..7027431 100644 --- a/2019-05-28_OCRUG_Hackathon_Captain_Rrrr/README.md +++ b/2019-05-28_OCRUG_Hackathon_Captain_Rrrr/README.md @@ -1,11 +1,14 @@ -# Best Visualization Award +# OCRUG 2019 Hackathon +## Best Visualization Award -## Team Name: Captain Rrrr +* 2019-05-28 +* Team Name: Captain Rrrr +* Team Members: Carey Huh, Crystal Yang, Dhruba Banerjee, Divya Swaminathan, Grace Gupta -### Team Members: Carey Huh, Crystal Yang, Dhruba Banerjee, Divya Swaminathan, Grace Gupta +## Abstract - Investigated the relationship between population size and water depletion - Found a sharper gradient (faster groundwater depletion) in agricultural vs. urban counties - Orange County showed an opposite relationship to nearby large urban counties that could not be explained - Water levels are highly correlated with drought severity -- Built a decision tree model to predict whether water level in 2019 is higher/lower compared to 2014 \ No newline at end of file +- Built a decision tree model to predict whether water level in 2019 is higher/lower compared to 2014 diff --git a/2019-05-28_OCRUG_Hackathon_Save_the_woRld/README.md b/2019-05-28_OCRUG_Hackathon_Save_the_woRld/README.md index bc2f53d..3efacc6 100644 --- a/2019-05-28_OCRUG_Hackathon_Save_the_woRld/README.md +++ b/2019-05-28_OCRUG_Hackathon_Save_the_woRld/README.md @@ -1,11 +1,14 @@ -# Best Model Award +# OCRUG 2019 Hackathon +## Best Model Award -## Team Name: Save The WoRld +* 2019-05-28 +* Team Name: Save The WoRld +* Team Members: Faizanul Syed Haque, Gregory Faletto, Javier Orraca, Sam Park, Shruhi Desai -### Team Members: Faizanul Syed Haque, Gregory Faletto, Javier Orraca, Sam Park, Shruhi Desai +## Abstract - Modelled the proportion of residents in fair/poor health in a county based on the contaminant levels and standard demographic controls - Investigated both ordinary least squares regression and generalized additive models - Arsenic appears to have a nonlinear effect not picked up by ordinary least squares - Started to look at how contaminants affect birth rates -- Found a general decline in health (across multiple measures) in California since 2010 \ No newline at end of file +- Found a general decline in health (across multiple measures) in California since 2010 diff --git a/2019-05-28_OCRUG_Hackathon_watR_inspectR/README.md b/2019-05-28_OCRUG_Hackathon_watR_inspectR/README.md index 68045d5..1a602a0 100644 --- a/2019-05-28_OCRUG_Hackathon_watR_inspectR/README.md +++ b/2019-05-28_OCRUG_Hackathon_watR_inspectR/README.md @@ -1,8 +1,11 @@ -# Best Insight Award +# OCRUG 2019 Hackathon +## Best Insight Award -## Team Name: watRinspectR +* 2019-05-28 +* Team Name: watRinspectR +* Team Members: Benjamin Lewis, Ethan Christensen, Leondra James, Mark Dela, Sarah Stewart -### Team Members: Benjamin Lewis, Ethan Christensen, Leondra James, Mark Dela, Sarah Stewart +## Abstract - Looked at the violations by type of violation, analyte type, county, and year - Looked at combinations of analyte and violation type in different counties in 2018 diff --git a/2019-06-25_Amazon_Alexa_Skill_Store/README.md b/2019-06-25_Amazon_Alexa_Skill_Store/README.md index d9e58fe..064145c 100644 --- a/2019-06-25_Amazon_Alexa_Skill_Store/README.md +++ b/2019-06-25_Amazon_Alexa_Skill_Store/README.md @@ -1,6 +1,8 @@ -#Speaker: Samira Kumar Vardharajan, Brandon Nguyen, Shruhi Desai, Qinyu Xi, Xinfang Zhang +# UCI MSBA Capstone Project +## Amazon Alexa Skill Store -##Title: Amazon Alexa Skill Store +* 2019-06-25 +* Speaker: Samira Kumar Vardharajan, Brandon Nguyen, Shruhi Desai, Qinyu Xi, Xinfang Zhang -###Abstract -Using effectively the skill intent model (a description of a third party application), including but not limited to, description, sample utterances, keywords, maturity rating, etc. we try to classify Alexa skills into one of 20 categories. Creating features with text data has a two-fold approach. The first is to use frequency of words to give words to each word occurring in the skill-intent model and use this matrix to predict the category as a supervised classification problem. The second is to use the semantic meaning, that is similarity of words, defined by cosine similarity. To overcome the big challenge of overlapping categories, we use a probabilistic approach to categorization. \ No newline at end of file +## Abstract +Using effectively the skill intent model (a description of a third party application), including but not limited to, description, sample utterances, keywords, maturity rating, etc. we try to classify Alexa skills into one of 20 categories. Creating features with text data has a two-fold approach. The first is to use frequency of words to give words to each word occurring in the skill-intent model and use this matrix to predict the category as a supervised classification problem. The second is to use the semantic meaning, that is similarity of words, defined by cosine similarity. To overcome the big challenge of overlapping categories, we use a probabilistic approach to categorization. diff --git a/2019-06-25_Pacific_Life_Predicincting_Policyholder_Lapse_Behavior/README.md b/2019-06-25_Pacific_Life_Predicincting_Policyholder_Lapse_Behavior/README.md index 424d5e6..fee9dcd 100644 --- a/2019-06-25_Pacific_Life_Predicincting_Policyholder_Lapse_Behavior/README.md +++ b/2019-06-25_Pacific_Life_Predicincting_Policyholder_Lapse_Behavior/README.md @@ -1,6 +1,8 @@ -#Speaker: Daniel Lee, Irene Yang, Pablo Barajas, Mason Hansen, Abhinav Kashyap  +# UCI MSBA Capstone Project +# Pacific Life: Predicting polcyholder lapse behavior -##Title: Pacific Life: Predicting polcyholder lapse behavior +* 2019-06-25 +* Speaker: Daniel Lee, Irene Yang, Pablo Barajas, Mason Hansen, Abhinav Kashyap  -###Abstract -Applying the principles of Statistics and Machine Learning to predict policyholder lapse rate for Variable Annuities. In addition, we gained insight into the underlying dynamics of policyholder lapse behavior. \ No newline at end of file +## Abstract +Applying the principles of Statistics and Machine Learning to predict policyholder lapse rate for Variable Annuities. In addition, we gained insight into the underlying dynamics of policyholder lapse behavior.