The document discusses how algorithms and big data are used to make predictions and recommendations. It provides examples of how data from sources like web searches, product sales, weather, and traffic can be analyzed to predict things like the spread of disease or demand during emergencies. Examples also discuss how insurance companies and Netflix use data to personalize services and recommendations. The document advocates for the benefits of algorithms and automation over human labor, noting that computers can process more data faster and without fatigue compared to people. It concludes with tips on how to create your own algorithm by defining the problem, outlining the manual process, and then automating it with a computer program.
Why algorithms are your future and how you can create one
1. Why algorithms are your
future and how you
can create one
@masterstips #DMS2015
2. Big data requires
multiple sources
• Population stats
• Impressions, traffic, sales
• Footfall
• Time of year/day
• Weather
• Customer data
• Previous year data
• The competition
@masterstips #DMS2015
3. Google flu
trends
Google’s flu predictor uses
search query volumes to track
the spread of influenza.
@masterstips #DMS2015
4. Big data to predict earthquakes
@masterstips #DMS2015
5. Hurricanes and
Pop-Tarts
Sales of Strawberry Pop-Tarts
increased by 7 times prior to a
hurricane.
The biggest pre-hurricane seller was
beer, not water.
@masterstips #DMS2015
9. Data insights
drive
Netflix can even use the data to
analyse the most popular elements
in movies and use this to create the
ultimate movie…
@masterstips #DMS2015
11. The more technology we invent, the
more big data we create.
@masterstips #DMS2015
12. The more big data we create, the more
we need bots to manage it.
@masterstips #DMS2015
13. “Every day, we create 2.5
quintillion bytes of data — so
much that 90% of the data in
the world today has been
created in the past two years
alone.” IBM
@masterstips #DMS2015
14. Car insurers are able to provide a
personalised quote in a few seconds
@masterstips #DMS2015
15. Businesses can track fleets
Black boxes in cars are
popular with insurance
companies.
@masterstips #DMS2015
17. Technology is coming
“In the world that we are creating very
quickly, we're going to see more and more
things that look like science fiction, and
fewer and fewer things that look like jobs.”
Andrew McAfee
@masterstips #DMS2015
18. Do not fear robots
taking your job.
Fear your failure
to let them.
@masterstips #DMS2015
19. Consider the humble traffic light
Lights automatically control traffic, 24 hours a day, all over the world.
@masterstips #DMS2015
21. Automation
Associated Press uses Automated
Insights software to automatically
write financial reports as stock
market data is released.
Financial reports
Computerisation has also reached
the area of financial advice.
@masterstips #DMS2015
22. A person can do one thing at a time
@masterstips #DMS2015
33. How to create an algorithm
1. Know what you want to know
2. Write down how you do it manually
3. Create a repetitive program that does the same
4. Put your feet up.
@masterstips #DMS2015
34. Get in touch
Steve Masters
@masterstips
@VerticalLeap
www.vertical-leap.uk
0845 123 2753
@masterstips #DMS2015
Notas del editor
Advanced Google search – PDF as search variable; include key phrases like “market reports” (use quote marks)
Mintel – UK consumer goods markets – mintel.com
ONS.gov.uk – manufacturing reports; business monitors
YouGov – may have some reports on general UK opinion and consumer trends
Management consultancy industry reports
Quandl.com – economics, finance & business data
FT.COM – register for free trial online subscription
Bloomberg.com
NY Times – NYTimes.com
EU provide lot of general information – PUBLICATIONS.EUROPA.EU
https://www.google.com/publicdata/
Economics Data
Useful Websites
American Economics Association RFE (Resources for Economics on the Internet – rfe.org). This site lists more than 2,000 resources available on the Internet of interest to academic and practicing economists, and those interested in economics.
Bized (bized.co.uk)
This site aimed at post 16 students is particularly strong in the provision of economics learning materials. Also explore the section on data handling skills. A good starting point if struggling with concepts.
Economic Data Freely Online – (http://www.economicsnetwork.ac.uk/links/data_free)
Excellent advice and links from John Sloman about free national, international and market data
ELDIS – (eldis.org)
A gateway to information on development issues, providing free and easy access to wide range of high quality online resources.
Gov.uk – (gov.uk)
Use this site to find the Government information from departments, publications or statistics in particular:
H.M. Treasury is useful for publications and statistics. The Office for Budget Responsibility provides forecasts and analysis of UK’s public finances
Each week, millions of users around the world search for health information online. As you might expect, there are more flu-related searches during flu season, more allergy-related searches during allergy season, and more sunburn-related searches during the summer. You can explore all of these phenomena using Google Insights for Search. But can search query trends provide the basis for an accurate, reliable model of real-world phenomena?
We have found a close relationship between how many people search for flu-related topics and how many people actually have flu symptoms. Of course, not every person who searches for "flu" is actually sick, but a pattern emerges when all the flu-related search queries are added together. We compared our query counts with traditional flu surveillance systems and found that many search queries tend to be popular exactly when flu season is happening. By counting how often we see these search queries, we can estimate how much flu is circulating in different countries and regions around the world. Ourresults have been published in the journal Nature.
http://www.world-earthquakes.com/index.php?option=sta
Animal behaviour
Changes in pressure or radon emissions
Electromagnetic variations
Precursor trends
Seismic gaps and patterns
http://www.nytimes.com/2004/11/14/business/yourmoney/14wal.html?_r=1&
HuRRICANE FRANCES was on its way, barreling across the Caribbean, threatening a direct hit on Florida's Atlantic coast. Residents made for higher ground, but far away, in Bentonville, Ark., executives atWal-Mart Stores decided that the situation offered a great opportunity for one of their newest data-driven weapons, something that the company calls predictive technology.
A week ahead of the storm's landfall, Linda M. Dillman, Wal-Mart's chief information officer, pressed her staff to come up with forecasts based on what had happened when Hurricane Charley struck several weeks earlier. Backed by the trillions of bytes' worth of shopper history that is stored in Wal-Mart's computer network, she felt that the company could "start predicting what's going to happen, instead of waiting for it to happen," as she put it.
The experts mined the data and found that the stores would indeed need certain products - and not just the usual flashlights. "We didn't know in the past that strawberry Pop-Tarts increase in sales, like seven times their normal sales rate, ahead of a hurricane," Ms. Dillman said in a recent interview. "And the pre-hurricane top-selling item was beer."
Amazon’s explanation of recommendations
http://www.amazon.com/gp/help/customer/display.html?nodeId=16465251
Amazon deploys tool to root out fake reviews
http://elliott.org/blog/amazon-deploys-new-weapon-war-phony-reviews/
Amazon’s algorithm will give extra weight to newer reviews, reviews from verified purchasers, and reviews that customers voted as being helpful. The change will also affect easy-to-spot 5-star ratings.
“The system will continue to learn which reviews are most helpful to customers and improve the experience over time,” Amazon spokeswoman Julie Law told ABC News.
The change started June 19 and may take some time to become apparent as Amazon’s new platform gradually alters star ratings and top reviews on product pages, CNET reported. Star ratings previously were averages of all reviews, so fake reviews could heavily influence the ratings.
Amazon’s latest attack on phony reviews follows its April filing of a lawsuit against the operators of three websites that it said sell fictitious praise for products. The operators denied the allegations.
https://labs.spotify.com/2014/02/28/how-to-shuffle-songs/
http://en.wikipedia.org/wiki/Fisher%E2%80%93Yates_shuffle
This was not random enough as it was causing clustering that made users think it wasn’t random.
They added an element of Floyd-Steinberg Dithering to create a more random result.
http://en.wikipedia.org/wiki/Floyd%E2%80%93Steinberg_dithering
There are 256 filters in the first layer of the network, which I numbered from 0 to 255. Note that this numbering is arbitrary, as they are unordered.
These four playlists were obtained by finding songs that maximally activate a given filter in the 30 seconds that were analyzed. I selected a few interesting looking filters from the first convolutional layer and computed the feature representations for each of these, and then searched for the maximal activations across the entire test set. Note that you should listen to the middle of the tracks to hear what the filters are picking up on, as this is the part of the audio signal that was analyzed.
http://benanne.github.io/2014/08/05/spotify-cnns.html
This data comes from everywhere: sensors used to gather climate information, posts to social media sites, digital pictures and videos, purchase transaction records, and cell phone GPS signals to name a few.
The term “big data” is also used in the context of data analysis. It’s not just about the existence of the data but also about the potential for computational analysis of that data. The data must be usable.
http://www.tracker.co.uk/
Also, Google collects big data related to drivers who use Google Maps – all of which helps to enhance the knowledge for the self-driving cars. This could also be used to manage traffic better and avoid congestion hotspots.
“In the world that we are creating very quickly, we're going to see more and more things that look like science fiction, and fewer and fewer things that look like jobs. Our cars are very quickly going to start driving themselves, which means we're going to need fewer truck drivers. We're going to hook Siri up to Watson and use that to automate a lot of the work that's currently done by customer service reps and troubleshooters and diagnosers, and we're already taking R2D2, painting him orange, and putting him to work carrying shelves around warehouses, which means we need a lot fewer people to be walking up and down those aisles.”
The first traffic control device appeared near the British House of Parliament at the intersection of George and Bridge Streets. The device was made in response to the desire by a Select Committee to use railway signals on highways. The device had lights and it used arms which extended outwards. It was operated manually by a police officer. The signal was 22 feet high and crowned with a gas light. The light was called the semaphore and had arms that would extend horizontally that commanded drivers to "Stop" and then the arms would lower to a 45 degrees angle to tell drivers to proceed with "Caution". At night a red light would command "Stop" and a green light would mean use "Caution".[7] The man behind this new and different invention was John Peake Knight a railroad engineer. The main reason for the traffic light was that there was an overflow of horse-drawn traffic over Westminster Bridge which forced thousands of pedestrians to walk next to the house of Parliament.[8] But Knight’s invention was not to last long. After only a month of use the device exploded and injured the police officer who was operating the light. In the first two decades of the 20th century semaphore traffic signals, like the one in London, were in use all over the United States with each state having its own design of the device. One good example was from Toledo, Ohio in 1908. The words “Stop” and “Go” were in white on a green background and the lights had red and green lenses illuminated by kerosene lamps for night travelers and the arms where eight feet above ground.[9] Controlled by atraffic officer who would blow a whistle before changing the commands on this signal to help alert travelers of the change, the design was also used in Philadelphia and Detroit.[10]The example in Ohio was the first time America tried to use a more visible form of traffic control that evolved the use of semaphore. The device that was used in Ohio was designed based on the use of railroad signals.[11]
The first three-colored signal and the traffic tower[edit]
In 1912, a traffic control device was placed on top a tower in Paris at the Rue Montmartre and Grande Boulevard. This tower signal was manned by a police woman and she operated a revolving four-sided metal box on top of a glass showcase where the word “Stop” was painted in red and the word “Go” painted in white. The traffic tower was featured in The Rider and Driver Magazine and soon the United States started to develop their own traffic control towers.[12] The tower was the first innovation that used the three-colored traffic signal and appeared first in the City of Detroit, where the first three-colored traffic light was built at the intersection of Michigan and Woodward Avenues in 1920. The man behind this three-color traffic light was police officer William Potts of Detroit. He was concerned about how police officers at four different lights signals could not change their lights all at the same time. The answer was a third light that was colored amber, which was the same color used on the railroad.[13] Potts also placed a timer with the light to help coordinate a four-way set of lights in the city. The traffic tower soon used twelve floodlights to control traffic and the reason for a tower in the first place was that at the time the intersection was one of the busiest in world, with over 20,000 vehicles daily.[14] The twelve-light system did not become available until 1928 and another feature of the light system was that hoods were placed over the light and each lens was sand-blasted to increase daytime visibility.[15] The popularity of the traffic tower, like the semaphore in the past decade, caught on in other states. Once again different states released their own design for the new traffic tower, and it was popular in places like New York and Detroit. In 1922 traffic towers were beginning to be controlled by automatic timers. The first company to add timers in traffic lights was Crouse Hinds. They built railroad signals and were the first company to place timers in traffic lights in Houston, which was their home city.[16] The main advantage for the use of the timer was that it saved cities money by replacing traffic officers. The city of New York was able to reassign all but 500 of its 6,000 officers working on the traffic squad; this saved the city $12,500,000.[17] The use of both the tower and the semaphores were done by 1930. The tower was great with starting innovation but they were too big and obstructed traffic and the semaphores were too small and drivers could not see them at night.[18]
Traffic light controlled by computer[edit]
The traffic light underwent many evolutions from a manually operated arm extending outwards in 1860s London to being controlled by a timer in a tower at an intersection. The control of traffic lights made a big turn with the rise of computers in America in the 1950s. Thanks to computers, the changing of lights made Crosby's flow even quicker thanks to computerized detection. A pressure plate was placed at intersections so once a car was on the plate computers would know that a car was waiting at the red light.[19] Some of this detection included knowing the number of waiting cars against the red light and the length of time waited by the first vehicle at the red.[20] One of the best historical examples of computerized control of lights was in Denver in 1952. One computer took control of 120 lights with six pressure-sensitive detectors measuring inbound and outbound traffic. The system was in place at the central business district, where the most traffic was between the downtown area and the north and northeastern parts of the city. The control room that housed the computer in charge of the system was in the basement of the City and County Building.[21] As computers started to evolve, traffic light control also improved and became easier. In 1967 the cities of Toronto and Ontario were the first to use more advanced computers that were better at vehicle detection.[22] Thanks to the new and better computers traffic flow moved even quicker than with the use of the tower. The computers maintained control over 159 signals in the cities through telephone lines. People praised the computers for their detection abilities. Thanks to detection computers could change the length of the green light based on the volume of waiting cars.[23] The rise of computers is the model of traffic control which is now used in the 21st century.
The first traffic control device appeared near the British House of Parliament at the intersection of George and Bridge Streets. The device was made in response to the desire by a Select Committee to use railway signals on highways. The device had lights and it used arms which extended outwards. It was operated manually by a police officer. The signal was 22 feet high and crowned with a gas light. The light was called the semaphore and had arms that would extend horizontally that commanded drivers to "Stop" and then the arms would lower to a 45 degrees angle to tell drivers to proceed with "Caution". At night a red light would command "Stop" and a green light would mean use "Caution".[7] The man behind this new and different invention was John Peake Knight a railroad engineer. The main reason for the traffic light was that there was an overflow of horse-drawn traffic over Westminster Bridge which forced thousands of pedestrians to walk next to the house of Parliament.[8] But Knight’s invention was not to last long. After only a month of use the device exploded and injured the police officer who was operating the light. In the first two decades of the 20th century semaphore traffic signals, like the one in London, were in use all over the United States with each state having its own design of the device. One good example was from Toledo, Ohio in 1908. The words “Stop” and “Go” were in white on a green background and the lights had red and green lenses illuminated by kerosene lamps for night travelers and the arms where eight feet above ground.[9] Controlled by atraffic officer who would blow a whistle before changing the commands on this signal to help alert travelers of the change, the design was also used in Philadelphia and Detroit.[10]The example in Ohio was the first time America tried to use a more visible form of traffic control that evolved the use of semaphore. The device that was used in Ohio was designed based on the use of railroad signals.[11]
The first three-colored signal and the traffic tower[edit]
In 1912, a traffic control device was placed on top a tower in Paris at the Rue Montmartre and Grande Boulevard. This tower signal was manned by a police woman and she operated a revolving four-sided metal box on top of a glass showcase where the word “Stop” was painted in red and the word “Go” painted in white. The traffic tower was featured in The Rider and Driver Magazine and soon the United States started to develop their own traffic control towers.[12] The tower was the first innovation that used the three-colored traffic signal and appeared first in the City of Detroit, where the first three-colored traffic light was built at the intersection of Michigan and Woodward Avenues in 1920. The man behind this three-color traffic light was police officer William Potts of Detroit. He was concerned about how police officers at four different lights signals could not change their lights all at the same time. The answer was a third light that was colored amber, which was the same color used on the railroad.[13] Potts also placed a timer with the light to help coordinate a four-way set of lights in the city. The traffic tower soon used twelve floodlights to control traffic and the reason for a tower in the first place was that at the time the intersection was one of the busiest in world, with over 20,000 vehicles daily.[14] The twelve-light system did not become available until 1928 and another feature of the light system was that hoods were placed over the light and each lens was sand-blasted to increase daytime visibility.[15] The popularity of the traffic tower, like the semaphore in the past decade, caught on in other states. Once again different states released their own design for the new traffic tower, and it was popular in places like New York and Detroit. In 1922 traffic towers were beginning to be controlled by automatic timers. The first company to add timers in traffic lights was Crouse Hinds. They built railroad signals and were the first company to place timers in traffic lights in Houston, which was their home city.[16] The main advantage for the use of the timer was that it saved cities money by replacing traffic officers. The city of New York was able to reassign all but 500 of its 6,000 officers working on the traffic squad; this saved the city $12,500,000.[17] The use of both the tower and the semaphores were done by 1930. The tower was great with starting innovation but they were too big and obstructed traffic and the semaphores were too small and drivers could not see them at night.[18]
Traffic light controlled by computer[edit]
The traffic light underwent many evolutions from a manually operated arm extending outwards in 1860s London to being controlled by a timer in a tower at an intersection. The control of traffic lights made a big turn with the rise of computers in America in the 1950s. Thanks to computers, the changing of lights made Crosby's flow even quicker thanks to computerized detection. A pressure plate was placed at intersections so once a car was on the plate computers would know that a car was waiting at the red light.[19] Some of this detection included knowing the number of waiting cars against the red light and the length of time waited by the first vehicle at the red.[20] One of the best historical examples of computerized control of lights was in Denver in 1952. One computer took control of 120 lights with six pressure-sensitive detectors measuring inbound and outbound traffic. The system was in place at the central business district, where the most traffic was between the downtown area and the north and northeastern parts of the city. The control room that housed the computer in charge of the system was in the basement of the City and County Building.[21] As computers started to evolve, traffic light control also improved and became easier. In 1967 the cities of Toronto and Ontario were the first to use more advanced computers that were better at vehicle detection.[22] Thanks to the new and better computers traffic flow moved even quicker than with the use of the tower. The computers maintained control over 159 signals in the cities through telephone lines. People praised the computers for their detection abilities. Thanks to detection computers could change the length of the green light based on the volume of waiting cars.[23] The rise of computers is the model of traffic control which is now used in the 21st century.