-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Spring 2017 7 years, 7 months ago
We are all drowning in data, and so is your future employer. Data pours in from sources as diverse as social media, customer loyalty programs, weather stations, smartphones, and credit card purchases. How can you […]
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 8 months ago
Just a reminder that your final exam will be on Friday, December 16 at 1:00pm in the same room as class. Please make sure that all missing assignments, quizzes and weekly questions are done before the start of […]
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 8 months ago
Here is the study guide for the third (final) exam.
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 8 months ago
Here is the link for the driver download
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 8 months ago
Here is the exercise.
And here is the spreadsheet you’ll need [In-Class Exercise 13.2 – VandelayOrdersAll.xlsx].
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 8 months ago
Here is the exercise
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 9 months ago
Here is the exercise.
And here is the spreadsheet you’ll need for the exercise [In-Class Exercise 12.2 – Sentiment Analysis Tools.xlsx].
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 9 months ago
Leave your response to the question below as a comment on this post by the beginning of class on November 30, 2016. It only needs to be three or four sentences.
What was the most important takeaway (from y […]
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 9 months ago
Leave your response as a comment on this post by the beginning of class on November 30, 2016. Remember, it only needs to be three or four sentences. For these weekly questions, I’m mainly interested in your op […]
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 9 months ago
Here is the exercise.
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 9 months ago
Some quick instructions:
You must complete the quiz by the start of class on November 28, 2016.
When you click on the link, you may see a Google sign in screen. Use your AccessNet ID and password to […] -
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 9 months ago
Here is the exercise.
Here is the excel spreadsheet you will need to complete this exercise [In-Class Exercise 11.2 – NCAA 2013-2014 Player Stats]
See Answer Key for part 6 here.
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 9 months ago
Leave your response as a comment on this post by the beginning of class on November 16, 2016. Remember, it only needs to be three or four sentences. For these weekly questions, I’m mainly interested in your o […]
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 9 months ago
Here is the exercise.
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 9 months ago
Some quick instructions:
You must complete the quiz by the start of class on November 14, 2016.
When you click on the link, you may see a Google sign in screen. Use your AccessNet ID and password to […] -
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 10 months ago
Here is the exercise.
And here is the Excel workbook you’ll need [Pew Story Data (Jan – May 2012).xlsx]
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 10 months ago
Leave your response as a comment on this post by the beginning of class on November 9, 2016. Remember, it only needs to be three or four sentences. For these weekly questions, I’m mainly interested in your op […]
-
The data in this article shows that household income has been increasing in multiple groups based upon age, race, household type, and regions. As a combined averaged they said it was an approximate 5.2% increase in all households as a total. Some groups such as the ages of 25-34 showed the highest increase in income which is hopeful data for all upcoming college graduates. One downside to the data provided is that the Northeast region had the lowest increase in income compared to the other regions but still showed a positive trend like the rest.
-
http://www.eia.gov/petroleum/weekly/
This article discusses about the merger and acquisitions of petroleum and other gasoline liquids in United States of America. I like the data in article provided because is it simple, to the point and most importantly easy to understand. Moreover the writer has aptly used different types of graphs such as bar charts , line graphs to show the different data. In order to contrast and compare the data, the writer made use of “Heart Rate” type of graph which makes the presented data easy to understand. -
http://www.rasmussenreports.com/public_content/politics/obama_administration/prez_track_nov4
This article contains data about the presidential polls. As it is so close to election day, I feel as if this data is very relevant to all of us. The election affects all us of immensely, so naturally we would be interest in data concerning the possible outcome. While this data is useful and of interest to many, it is notoriously unreliable, as it relies on people telling the truth. However, the data is still interesting to see, as it gives us an idea about the current political climate in the US.
-
http://fivethirtyeight.com/features/has-trump-already-lost-nevada/
This article discusses how Trump will most likely already lose Nevada. According to polls on 538, Trump has just recently won Nevada, but due to early voting, they’re suspecting their predictions were off. Early voters are already favoring Clinton. If Trump loses Nevada, he most likely wont have a shot at winning this election. However, according to 538, most states are seeing an incline in Trump support; with a few days left, this could make an extremely close election. If Trump were to win, he would most likely lose the popular vote, but win the electoral college. -
http://www.wsj.com/articles/early-voting-data-shows-whos-turning-out-1478281004
This article discusses updates on the current presidential election. Due to early and absentee voting, data has already begun collecting regarding the election. The article includes some of this data that has been gathered thanks to these early voters. The article discusses the majority voters in the early voting, as well has who has the advantage so far in the election. So far, the turn out is not exactly how many expected, but it is hard to tell based solely on early voting, only the official election will reveal the winner. -
http://fivethirtyeight.com/features/how-much-did-comey-hurt-clintons-chances/
In this article it discusses the current percentages of Hilary’s Clintons chances of becoming elected for President. Clinton had an 81 percent chance of winning the election on October 28, according to the websites polls forecast. Today, her chances are 65 percent, according to the same forecast. The changes have dipped down by the popular vote. This article is interesting to me to see how the public can can be persuaded still to change there vote, even after knowing Donald Trump will use the same “Email” tactic against her. Data is used in this article in the form of percentages, and research conducted through out the two parties. -
http://fivethirtyeight.com/features/doctor-strange-magic-superhero-movies/
This data shows that superhero movies are becoming more popular than regular fiction with magical elements. I find this data interesting because I am a Film major and it is important to know what features will be popular before investing time and money into producing them. -
http://fivethirtyeight.com/features/in-an-election-defined-by-race-how-do-we-define-race/
The election has been very defined by race, with Donald Trump’s supporters being majority white. The article and the data itself is used to explain what we determine as race. It’s interesting to see that sometimes even a fraction of a race in your bloodline can determine what you’re determined to be.
-
http://fivethirtyeight.com/features/the-senates-a-toss-up-a-race-by-race-look/
This article looks at races going on in the U.S. Senate on Election Day. It breaks down the races from a national and state perspective, explaining how Clinton’s downturn in the polls last week has hurt Senate Democrats. The polls are close, and the outcomes of the races could determine which party takes control of the Senate. What’s even more interesting about the article, is the fact that it breaks down which Senate races are toss-ups, and which ones lean to the Democrat or Republican candidate running. Predicting the outcome of the election has been driven entirely by data, and it’s interesting to see how the data breaks down in races other than the presidential one.
-
http://fivethirtyeight.com/features/the-broncos-pass-defense-is-somehow-even-better-this-season/
This article spoke about the Denver Broncos’ pass defense. This article was remarkable because after having the number pass defense later year, and winning the Superbowl, it is said, on paper, that this years’ defense is better. They used net yards per pass attempt to measure the stats. The Broncos are -1.11 yards/game better than the rest of the NFL average. I like this because i am a big football guy, espically with fantasy football, and information like this helps when deciding which team to bet on, and which defense to use each week in my fantasy lineups.
-
http://www.aol.com/article/news/2016/11/07/how-many-people-voted-in-the-last-presidential-election/21600631/
If you scroll down to the chart that lets you choose what state you can look at specifically it will generate the data of the percentage of people who voted for which candidate in that state. This interest me to see the differences in some states such as our own and how close each candidate came to one another. Pennsylvania was one of the closest calls bringing in 48% for Clinton and 49% for Trump. Our state is a prime example that every vote can count and could have swayed this either way. On the other hand there were some drastic differences between the candidates in other areas such as the District of Columbia that proceeded to show 93% Clinton and only 4% Trump. -
https://www.wired.com/2016/11/trump-polling-data/
In this article it discusses how Trump’s political team was able to predict that he had a solid chance at winning the presidency. His team were able to determine his chances by analyzing the polling data that came in during early voting times. They took that data and reconstructed their prediction model which allowed them to more accurately predict how the election could turn out. Furthermore, the data they evaluated helped them determine which undecided states Trump could possibly win.
-
http://fivethirtyeight.com/features/final-election-update-theres-a-wide-range-of-outcomes-and-most-of-them-come-up-clinton/
In this article it summarizes the outcome of the election 2016. It shows several graphs on what people predicted would be the outcome, most predictions predicted that Hilary Clinton would win; however these proved to be wrong leaving Donald Trump the president. It gives an example of many ways Hilary Clinton could have won if state polls were slightly off.It proves that people thought Clinton had around 78% chance of winning while Trump had around only 28% chance. It also goes into explanation on how many undecided voters there were and there were a lot more than previous years. This proved that estimates and predictions may no longer be trusted and we should realize the impact on our vote. -
http://fivethirtyeight.com/features/immigrants-are-keeping-america-young-and-the-economy-growing/
This article highlights the importance of immigrants in America in light of this year’s election. Immigrants have a higher labor-force participation rate than native-born Americans. These people are more likely to start businesses and create more jobs for everyone. The article claims that, “on average”, everyone is better off as a result of immigrants. However, there will always be people on the extremes of either side of the argument.
-
http://projects.fivethirtyeight.com/2016-election-forecast/?ex_cid=rrpromo
This article highlights the likelihood of who would win the 2016 Election for President of the United States. It was interesting to see the data-centric predictions that ended up being completely wrong. Particularly, I believe that the best visualization used was ‘the winding path to 270 electoral votes’. This allowed the viewer to better understand how likely it actually might be given the predictions, even though the path went much different on 11/8. -
The data in this article looks at the rise and spread of obesity in each state. This data was collected for both men and women over the past three decades. As of today two-thirds of adults in the United States fall into the category of overweight or obese. From the data there is a higher obesity rate in southern states then across he country. In 2015 a majority of states have an obesity rate of over 30 %. This is interesting because this can show where our country is headed concerning obesity and can allow people to predict how high the obesity rate will be in the near future. -
https://fivethirtyeight.com/features/can-the-warriors-break-basketball-again/
This article talks about the Golden State Warriors potential to set the record for most wins in a regular season NBA basketball for two seasons in a row. Last season the Warriors broke the NBA record set by the 95-96 bulls by winning 73 games and only accumulating 9 losses during the regular season. The Warriors went ahead this summer and acquired superstar and 5th best ranked NBA player Kevin Durant to their roster, pairing him with best statistical player in the league Stephen Curry. The warriors last season were the best team in shooting the basketball in NBA history and now they have added one of the leagues premier shooters without having many demising losses from last seasons roster. The author of this article created a model to show just how much better the Warriors are shooting than everybody else. It is “I brewed my own model for expected points for a given shot based on its location, the position and height of the nearest defender, how many dribbles the player took, how long he held the ball for, and how much time was left on the shot clock. Using this, we can calculate points contributed relative to expectation for each shot taken and thus how much value players added from shooting ability alone.” Based on his calculations of points per shot against expectations, Curry and Durant were the 2 most efficient shooters in the 2015-2016 NBA season Durant was expected to Score .98 points per shot and Curry 1.01. These numbers may seem low however you must factor in these players shot many more times than practically every other athlete in the NBA. These two athletes along with 3rd player on the roster Klay Thompson are all outliers in NBA statistical shooting categories, while they also perform very efficiently on the defensive side of the ball. On open 3 pointers Curry Thompson and Durant all made over 40 % of their attempts and converted on 35 % of their contested 3 point attempts 2 very impressive stats. The warriors now appear to have a monopoly in shooting from 3 point range. This article is very interesting to me because I am a huge NBA fan, and the idea that the greatest NBA team last season became better in the off season is quite impressive to me. This is relevant to me because I aspire one day to work for a sports organization in their data analytics department working with any type of statistics. Overall with this addition to their team I do believe it is very likely for the 2016-2017 Warriors to outperform their predecessor. -
http://projects.fivethirtyeight.com/2016-election-forecast/
This article talks about the chances that each presidential candidate has of winning the election. It shows that Hillary Clinton was most likely to win the presidency. This article also features many different in-depth charts that show exactly how each state was projected to vote as well as what states could be tipping points. This is especially interesting on a day like today where you can see how wrong this data was and how the entire election actually tipped in the other direction. -
http://www.pcworld.com/article/3140175/big-data/is-trumps-unexpected-victory-a-failure-for-big-data-not-really.html
This article talks about how the polling for the presidential election had predicted Hillary Clinton to win the election, and what the problem was with the data as Donald Trump won the election. The failed predictions could shine light on the doubts of big data. Forecasters like FiveThirtyEight have said that the problem with the polls has more to do with the data collection than the number crunching. Professor Samuel Wang states that it might have been a problem with polls undercounting hard-to-reach voters. In other cases, some voters may have lied to the polls. -
http://www.cnn.com/election/results
This article shows all the results from last nights election and who won in the different states. It contains a lot of data on the demographics on which gender voted for each candidate and also the different age demographics that voted for either one. It is relevant to me because it helps me to see which states voted for who and also which ages and genders so I can see how i fit in with the rest of the people in our country. It is interesting because it gives us a lot of incite into the other parts of our country and how we expect things to turn out in future elections -
http://projects.fivethirtyeight.com/2016-election-forecast/?ex_cid=rrpromo
This article is from fivethirtyeight.com and shows the predictions for the presidential election. It’s a nice and simple map that shows the chances of the politician winning and what their chance of winning each state is. I used this a lot around the election to see who had the upper hand in it. For about 99% of the time it said Hillary was going to win but they were wrong and Trump took the election. But it was still fun to see how each state would lean and to see who has the biggest chance of winning/ -
http://phillyelectionresults.com
This website gives the election results for voters in Philadelphia and breaks down the candidates by the percentage of votes that they received. Other than that, there isn’t much data that was provided. However, it does show that all Democrats won in the Philadelphia election, showing the more liberal atmosphere that Philadelphia has. -
http://www.msn.com/en-us/news/politics?ocid=spartanntp
This is an interactive map that shows the results of the election state by state and how many electoral votes were won by the candidate. Also, a percentile of how mush each candidate won the state by. I find this interesting because it helps a voter or observer see what state was won by whom and by how much. -
http://fivethirtyeight.com/features/the-undocumented-immigrant-workforce-isnt-growing/
The article that i have chosen talks about how the immigrant workforce is not actually growing as most people think. This article interested me because for the last few months Donald Trump has made immigration one of his main speaking points and many Americans feel as though the immigrants are taking all the jobs. Since 2006, the immigrant workforce has actually plateaued at around 8 million. Only 17% of agricultural jobs and 13% of construction jobs are made up of immigrants. This shows that one of the biggest fears of Americans is actually just illusion. -
we just went through with the us elections and saw there were major discrepancies in public polls and the results. To tackle this statisticians in the upcoming election have a created a new model called hidden markov mode.The way it works is that the two-party preferred national voting intention is a hidden latent variable that is modelled everyday in the period of analysis. Election results provide anchors, which allow for polls to predict national voting intentions while simultaneously taking into account of systematic in-house bias and sample size of polls. I think it would really help not the parties but the nation if the polls were right on making the right decision.
-
http://fivethirtyeight.com/features/gun-deaths/
This article contains data about gun deaths in America and is also an interactive graphic which lets the user explore by categories the amount of the deaths caused by guns from types (suicides/homicides/accidents/undetermined), gender (male/female), age (all, under 15, 15-34, 35-64, 65 and older), and also race (White, Black, Asian, Hispanic, and Native American). It is really interesting because we can tell how these deaths are caused by. -
In this article it summarizes the outcome of the election 2016. It depicts several visualizations on what people predicted would be the outcome, most predictions insinuated Hilary Clinton would be the victor; however these proved to be wrong as Donald Trump ultimately became the president. It gives an example of many ways Hilary Clinton could have won if state polls were slightly off. It displays that Americans thought Clinton had around 77% chance of winning while Trump had around only 28% chance. It also goes into explanation on how many undecided voters there were and there were many more than previous elections.
http://fivethirtyeight.com/features/final-election-update-theres-a-wide-range-of-outcomes-and-most-of-them-come-up-clinton/ -
http://fivethirtyeight.com/features/atts-merger-could-be-a-bad-sign-for-the-economy/
The article above talks about how the AT & T and Time Warner merger poses a serious threat to the health of our economy. In more recent years we’ve noticed trends that indicate we aren’t creating new companies and the amount of existing startups is quickly diminishing. With a Merger of two incredibly large companies such as these we could see a potential monopoly develop which would allow for easy price hikes from the company and possible preferrential treatment. Overall we’re seeing the economy becoming much less flexible, and with the addition of such a big merger it is quite a possibility that it will become harder for startups to succeed thus limiting innovation. -
http://hotair.com/archives/2016/11/08/early-voting-data-in-north-carolina-encouraging-for-trump/ This article shows how the data projected election results in North Carolina. This was a hugely important state for Trump to win if he wanted a chance at claiming the election. As we know, he won North Carolina and ultimately the election. This shows demographic data for ages as well as political and social affiliations.
-
How Netflix’s Movie Library Compares to Amazon Prime and HBO Now
The article above describes how Netfix, despite its shrinking library, has the best overall quality compared to hbo and amazon prime. This is relevant to me because it shows how competitive the market for streaming entertainment has become. As netflix invests less money in retaining old things in their library, they focus more on creating new, high quality content. Whether this is a good thing, I don’t know,
-
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 10 months ago
Here is the exercise.
And here are the workbooks [2012 Presidential Election Results by District.xlsx and Portrait 113th Congress.xlsx]
-
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 10 months ago
Some quick instructions:
You must complete the quiz by the start of class on November 7, 2016.
When you click on the link, you may see a Google sign in screen. Use your AccessNet ID and password to […] -
Ermira Zifla wrote a new post on the site MIS 0855: Data Science Fall 2016 7 years, 10 months ago
Here are the assignment instructions. Groups MUST be 4 or 5 members. You may not do this assignment on your own or in smaller groups than 4.
The assignment is due December 5, 2016. We’ll do the pre […]
- Load More