Experts say that up to 26 tonnes of globally fishing is illegal, unreported, or unregulated. Sadly, this is equal to a third of all fish that are caught legally.

Illegal fishing may be conducted with prohibited gear and result in overfishing, resulting in damage to coral reefs, seamounts, and other vulnerable marine ecosystems. 55% of the world’s reefs are threatened by excessive and destructive fishing. In South Asia, nearly 95% of all reefs are affected, and it is believed that most of the remote coral reefs in the world are overfished.

Other times, illegal fishing may result in the capture of protected species like the Atlantic salmon, which is endangered, or non-target species like sharks and turtles.

Fishing is an incredibly difficult industry to monitor, but some organisations are leveraging AI to tackle this problem. Last year, The Nature Conservancy instigated a challenge on Kaggle, a data science competition platform, to source a model that could detect when protected fish were caught. The software would then flag the portions of the video suspect of illegal fishing to be reviewed by a human. It currently takes 6 hours for an officer to review 10 hours of fishing footage, so this technology would increase the scrutiny placed on fishermen.



How AI will save fishing

Beekeeping and AI 
Bees are an integral part of the world's overall ecosystem and even for agriculture which is a sourcer of food for us all.

The bees are dying, the bees are dying! Colony collapse disorder is when a majority of worker bees in a colony die, leaving the queen bee alone in the hive without sustenance. This has been a massive problem in recent years — in 2016 beekeepers in the US lost 44% of their colonies. Since most plants rely heavily on bees for pollination, their absence is strongly felt throughout the entire ecosystem. There are many speculations about why colony collapse disorder is happening, such as overuse of pesticides, but to this day no one is quite sure what the reason is.

Even if we don’t know how to prevent bee colonies from collapsing, we can still focus on keeping the ones we have healthy! Startups like Nectar, based out of Montreal, are developing IoT solutions powered by AI to help beekeepers maintain healthy colonies. Nectar is building an apiary management platform, so that beekeepers can use data rather than intuition to determine what bees need and when they need it. The technology is even able to predict what events will occur in a hive and when, depending on many factors such as the temperature and season.

The majority of bees in the world have died out.

Bee-keeping Will Keep the Earth Alive

Rainforests are an important home for thousands of species as well as helping towards global warming.
Rainforests contain half of all life on Earth’s land mass and produce 20% of our oxygen, despite only covering 2% of our planet’s surface. However, they are disappearing at an alarming rate. It is estimated that up to 90% of all deforestation in tropical rainforests is illegal, wreaking havoc on the diverse animal and plant life that call the rainforest their home. One of the biggest drivers for this has been agriculture, as cattle ranchers transform the land to be used to raising livestock. Sadly, deforestation is difficult to detect and about a fifth of the Amazon rainforest has been lost in the last 50 years.

One of the reasons that it’s difficult to protect the rainforest is its size. The Amazon Rainforest is 23 times larger than the United Kingdom, making it impossible to monitor manually. It’s for this reason that organizations like The Rainforest Connection are turning to AI. The Rainforest Connection has combined solar panels and cells phones to create a cheap and rigorous acoustic monitoring system. They call these Guardian devices, and use AI to monitor for sounds of logging, such as chainsaws and vehicles, that are far away and imperceptible to the human ear. When these sounds are detected, the device alerts authorities in real time to intervene.

AI preventing deforestation

AI's application in terms of agriculture. Due to the varying factors that affect soil quality, and the dessertification of several parts of the world's previously fertile land, AI is helping to improve soil management:
Soil is the foundation of agriculture, without clean and healthy earth all we are left with is dirt. Much of the soil found on earth is unusable or inaccessible since it’s found underneath bodies of water (the ocean, rivers, lakes), therefore the formation of quality soil is critical for agriculture. There are multiple factors that can influence the quality of soil, namely parent material, climate, location, the chemicals and substances found in water, human and animal interaction, etc. It feels like you almost need fate, or luck on your side for all of these elements to align, but then there is also the time factor. It takes 500 years for an INCH of topsoil to form.

Increased human activity has led to urban development pollution, landfills and climate change, causing the amount of farmable land to decrease despite population being at an all time high. Companies like Trace Genomics have developed technology that can improve soil management. The process is fairly simple: you collect a sample of your soil and send it in for genome sequencing, computational analysis, and data science to analyse and measure your soil health. Afterwards, Trace Genomics provides you with custom recommendations for how you can improve your soil health and crop yield.

AI and Agriculture

You might have seen initiatives like aqua drones and jelly fish filters cleaning our oceans, or desalination and wastewater treatment technologies being used to help curb the freshwater crisis.

One area where AI is making vast improvements is in the development of smarter cities. Fracta is a company that uses machine learning to plan water main replacements. In order to avoid something they call “pipeaggedon” they are analyzing large data sets in order to shed light on what’s really happening with a city’s infrastructure. The technology can also accurately predict the possibility of water main failure for the fraction of the cost in a fraction of the time. Not only does benefit the city itself, but the taxpayer who would suffer the consequences of dirty (or lack of) water and ultimately be responsible for the cost.

AI and water preservation

Here are some great organizations working on AI to save the planet

They have big partnerships with Microsoft and such so it's pretty cool.

Organizations to watch in the future of Green AI

https://www.stradigi.ai/blog/thinking-green-ai-for-earth/

Technology has been enemy #1 to most environmentalists for a number of reasons: the manufacturing industry creates a tremendous amount of dangerous waste, machinery technology has increased development projects which has led to the extinction of many plants, birds and animals and our excessive power consumption is one of the leading causes of greenhouse gas emissions.

It isn’t all doom and gloom though, because as it turns out, AI has a green thumb and it’s been used to develop technology that can help prevent and in some cases even restore the damage we have done.

If human beings continue to refine technology and make a serious stand in protecting the environment, AI will become the planet’s biggest ally in our goal to become a more eco friendly planet.

University of Michigan - Ann Arbor

This node can be used to share links to online articles and websites that are beneficial for learning and are references for other nodes. 

Helpful Data Science Online Articles

 [statisticsbyjim.com](https://statisticsbyjim.com/) 

Helpful Statistics Blog 

Assumptions Of Linear Regression Algorithm. Gomathi Tamilselvam. (2019, May 24). Retrieved from [towardsdatascience.com/assumptions-of-linear-regression-algorithm-ed9ea32224e1](https://towardsdatascience.com/assumptions-of-linear-regression-algorithm-ed9ea32224e1) 

Assumptions Of Linear Regression Algorithm and Scikit-learn Visualizations

https://www.r-bloggers.com/how-to-detect-heteroscedasticity-and-rectify-it/

Test for Heteroscadasticity in R

LOGISTIC REGRESSION. Maheshwari, Vineet. (2018, December). Retrieved from
[medium.com/datadriveninvestor/logistic-regression-18afd48779ce](https://medium.com/datadriveninvestor/logistic-regression-18afd48779ce) 

Logistic Regression

How Good is Your Model? — Intro to Resampling Methods. Peixeiro, Marco. (2019, January). Retrieved from [towardsdatascience.com/how-good-is-your-model-intro-to-resampling-methods-2831e832042a](https://towardsdatascience.com/how-good-is-your-model-intro-to-resampling-methods-2831e832042a) 

Intro to Resampling Methods

Dhiraj, K. (2019, August 29). Difference between K-Nearest Neighbor(K-NN) and K-Means Clustering. Medium. [medium.com/@dhiraj8899/difference-between-k-nearest-neighbor-k-nn-and-k-means-clustering-d9a44859182f](https://medium.com/@dhiraj8899/difference-between-k-nearest-neighbor-k-nn-and-k-means-clustering-d9a44859182f) 

Medium: Difference between K-Means and KNN

Lagrangians are a tool for constrained optimization problems. Have a look at the video for an intuitive explanation: 

[www.khanacademy.org/math/multivariable-calculus/applications-of-multivariable-derivatives/lagrange-multipliers-and-constrained-optimization/v/the-lagrangian](https://www.khanacademy.org/math/multivariable-calculus/applications-of-multivariable-derivatives/lagrange-multipliers-and-constrained-optimization/v/the-lagrangian) 

Lagrangian Khan Academy Video

Data Set. Wikipedia. (2020, February 17). Retrieved from [en.wikipedia.org/wiki/Data_set](https://en.wikipedia.org/wiki/Data_set) 

Data Set

Amir Azimi. Is ‘Computer Science’ and ‘Information Systems’ the same subject? (2017, May 9). Retrieved from [medium.com/@HazzAzimi/is-computer-science-and-information-systems-the-same-subject-4e214292426e](https://medium.com/@HazzAzimi/is-computer-science-and-information-systems-the-same-subject-4e214292426e) 

Is ‘Computer Science’ and ‘Information Systems’ the same subject?

DelSole, Michael. (2018, April 24). *What is One Hot Encoding and How to Do It*. Medium. Retrieved from [medium.com/@michaeldelsole/what-is-one-hot-encoding-and-how-to-do-it-f0ae272f1179](https://medium.com/@michaeldelsole/what-is-one-hot-encoding-and-how-to-do-it-f0ae272f1179) 

What is One Hot Encoding and How to Do It

https://www.dataquest.io/blog/15-python-libraries-for-data-science/

Python Libraries for Data Science

Ph.D., Benjamin Obi Tayo. “Data Scientist Coding Exercise.” Medium, Towards AI - Multidisciplinary Science Journal, 11 June 2020. Retrieved from:  

https://medium.com/towards-artificial-intelligence/data-scientist-coding-exercise-e62f4de7df9e  



Data Scientist Coding Excercise

https://realpython.com/invalid-syntax-python/

Invalid Syntax in Python: Common Reasons for SyntaxError

Matplotlib Documentations

The Big Data Landscape is an annual article written by VC Matt Turck that provides an extremely granular view of companies within the Data Science space along with high and lowlights of the year. 

The Big Data Landscape can be accessed here: http://mattturck.com/wp-content/uploads/2019/07/2019_Matt_Turck_Big_Data_Landscape_Final_Fullsize.png



The Big Data Landscape of 2019

Sætra, Henrik Skaug. “Science as a Vocation in the Era of Big Data: The Philosophy of Science behind Big Data and Humanity’s Continued Part in Science.” Integrative Psychological & Behavioral Science, vol. 52, no. 4, Springer Science and Business Media LLC, 2018, pp. 508–22, doi:10.1007/s12124-018-9447-5.

https://search.lib.umich.edu/articles/record/FETCH-LOGICAL-14812-c1be5f7dc43257eb09516e3b08ea65e044c89b4fb42722c847b69ff64de0a37a3

Science as a Vocation in the Era of Big Data: the Philosophy of Science behind Big Data and humanity’s Continued Part in Science

Hicks, Stephanie C, and Irizarry, Rafael A. “A Guide to Teaching Data Science.” The American Statistician, vol. 72, no. 4, Informa UK Limited, 2017, pp. 382–91, doi:10.1080/00031305.2017.1356747.

https://search.lib.umich.edu/articles/record/FETCH-LOGICAL-14293-bb9dfa170b88df65690f79b02776e19ee338a692c0f0dcb1cc77fe06383a319c3

A Guide to Teaching Data Science

Van Horn, John, et al. “Advancing the International Data Science Workforce through Shared Training and Education.” F1000 Research, vol. 8, F1000 ( Faculty of 1000 Ltd), 2019, p. 251, doi:10.12688/f1000research.18357.1.

https://search.lib.umich.edu/articles/record/FETCH-LOGICAL-12021-e001cfa097b5c4e10c0722b63c521b8f3bc87c6c4a2fbc69395ee643b76403783

Advancing the international data science workforce through shared training and education


https://www.datamation.com/big-data/structured-vs-unstructured-data.html

Structured Data vs. Unstructured Data

An interesting article that looks at Baye's theorem from three different perspective: https://slackprop.wordpress.com/2016/08/28/the-three-faces-of-bayes/

The Three Faces of Bayes

https://www.unofficialgoogledatascience.com/2016/10/practical-advice-for-analysis-of-large.html

Practical advice for analysis of large, complex data sets

"Decision intelligence is the discipline of turning information into better actions at any scale." This article explores this idea: https://towardsdatascience.com/introduction-to-decision-intelligence-5d147ddab767

Introduction to Decision Intelligence

How Artificial Intelligence will save the environment

Business intelligence (BI) systems have steadily grown faster, simpler to use, more accurate at processing massive amounts of internal and external data, as well as identifying patterns that were previously undetected, or even undetectable, by human users. And yet, some of the more popular BI systems on the market do little more than repackage and distil existing data about past performance, albeit in a visually palatable format. A new approach to mining and deciphering unstructured data using algorithms and automated machine learning (AutoML) to forecast changes in the business landscape before they happen has quite rightly been hailed as the next big thing in artificial intelligence. Here’s how next-generation business platforms that leverage these predictive capabilities put you in a better position than traditional BI.

https://www.stradigi.ai/blog/5-ways-advanced-predictive-analytics-outshine-bi/

How advanced predictive analytics is outpacing old school business intelligence

Maintaining Data Privacy in Association Rule Mining. (2002, August). Retrieved from https://www.sciencedirect.com/science/article/pii/B9781558608696500664

Maintaining Data Privacy in Association Rule Mining

Nonnegative Matrix Factorization: An Analytical and Interpretive Tool in Computational Biology. (2008, July) Retrieved from
https://www.researchgate.net/publication/51427894_Nonnegative_Matrix_Factorization_An_Analytical_and_Interpretive_Tool_in_Computational_Biology

Nonnegative Matrix Factorization

Esbroeck, A. & Chia, C.-C & Syed, Z.. (2012). Heart rate topic models. 2. 1635-1641.  Retrieved from https://www.researchgate.net/publication/290032961_Heart_rate_topic_models

Heart Rate Topic Models

Daniel Gruhl, R. Guha, Ravi Kumar, Jasmine Novak, and Andrew Tomkins. 2005. The predictive power of online chatter. In Proceedings of the eleventh ACM SIGKDD international conference on Knowledge discovery in data mining (KDD '05). Association for Computing Machinery, New York, NY, USA, 78–87. DOI:https://doi.org/10.1145/1081870.1081883

The Predictive Power of Online Chatter

Ver Steeg, Greg & Galstyan, Aram. (2011). Information Transfer in Social Media. 10.1145/2187836.2187906. Retrieved from https://www.researchgate.net/publication/51944496_Information_Transfer_in_Social_Media

Information Transfer in Social Media

Danescu-Niculescu-Mizil, Cristian & Sudhof, Moritz & Jurafsky, Dan & Leskovec, Jure & Potts, Christopher. (2013). A computational approach to politeness with application to social factors. ACL 2013 - 51st Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference. 1.  Retrieved from  https://www.researchgate.net/publication/239522833_A_computational_approach_to_politeness_with_application_to_social_factors

A computational approach to politeness with application to social factors

Aniruddha Bhandari, (2020). AUC-ROC Curve in Machine Learning Clearly Explained. Retrieved from https://www.analyticsvidhya.com/blog/2020/06/auc-roc-curve-machine-learning/

AUC-ROC Curve in Machine Learning

Ritter, Alan & Mausam, Mausam & Etzioni, Oren & Clark, Sam. (2012). Open domain event extraction from Twitter. Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 1104-1112. 10.1145/2339530.2339704. Retrieved from https://www.researchgate.net/publication/254464301_Open_domain_event_extraction_from_Twitter

Open Domain Event Extraction from Twitter

Gui, Tao & Huang, Haoran & Peng, Minlong & Huang, Xuanjing. (2017). Part-of-Speech Tagging for Twitter with Adversarial Neural Networks. 2411-2420. 10.18653/v1/D17-1256. Retrieved from https://www.researchgate.net/publication/322587060_Part-of-Speech_Tagging_for_Twitter_with_Adversarial_Neural_Networks

Part-of-Speech Tagging for Twitter with Adversarial Neural Networks

Derczynski, Leon & Ritter, Alan & Clark, Sam & Bontcheva, Kalina. (2013). Twitter Part-of-Speech Tagging for All: Overcoming Sparse and Noisy Data. International Conference Recent Advances in Natural Language Processing, RANLP. Retrieved from https://www.aclweb.org/anthology/R13-1026.pdf

Twitter Part-of-Speech Tagging for All: Overcoming Sparse and Noisy Data

https://towardsdatascience.com/probability-and-statistics-for-data-science-part-1-3eed6051c40d

Probability and Statistics for Data Science

https://towardsdatascience.com/how-fast-numpy-really-is-e9111df44347gi=2448916cc6c0#:~:text=As%20the%20array%20size%20increase,also%20frees%20the%20memory%20faster

How Fast Numpy Really is and Why?

https://towardsdatascience.com/types-of-data-sets-in-data-science-data-mining-machine-learning-eb47c80af7a

Types of Data Sets in Data Science, Data Mining & Machine Learning

Kong, Lingpeng & Schneider, Nathan & Swayamdipta, Swabha & Bhatia, Archna & Dyer, Chris & Smith, Noah. (2014). A Dependency Parser for Tweets. 10.3115/v1/D14-1108. Retrieved from https://www.aclweb.org/anthology/D14-1108.pdf

A Dependency Parser for Tweets

Hale, John & Kuncoro, Adhiguna & Hall, Keith & Dyer, Chris & Brennan, Jonathan. (2019). Text Genre and Training Data Size in Human-like Parsing. 5849-5855. 10.18653/v1/D19-1594.  Retrieved from https://www.aclweb.org/anthology/D19-1594.pdf

Text Genre and Training Data Size in Human-Like Parsing

Röder, Michael & Both, Andreas & Hinneburg, Alexander. (2015). Exploring the Space of Topic Coherence Measures. WSDM 2015 - Proceedings of the 8th ACM International Conference on Web Search and Data Mining. 399-408. 10.1145/2684822.2685324. 

Exploring the Space of Topic Coherence Measures

Derczynski, Leon & Chester, Sean. (2016). Generalised Brown Clustering and Roll-up Feature Generation. 

Generalised Brown Clustering and Roll-Up Feature Generation

https://machinelearning.wtf/terms/symmetry-breaking/#:~:text=Symmetry%20breaking%20refer%20to%20a,This%20is%20the%20%E2%80%9Csymmetry%E2%80%9D.

Symmetry breaking


Thắng, Lương & Sutskever, Ilya & Le, Quoc & Vinyals, Oriol & Zaremba, Wojciech. (2014). Addressing the Rare Word Problem in Neural Machine Translation. 10.3115/v1/P15-1002. Retrieved from https://arxiv.org/pdf/1410.8206.pdf

Addressing the Rare Word Problem in Neural Machine Translation

Weston, Jason & Dinan, Emily & Miller, Alexander. (2018). Retrieve and Refine: Improved Sequence Generation Models For Dialogue. 87-92. 10.18653/v1/W18-5713. Retrieved from https://arxiv.org/pdf/1808.04776.pdf.

Retrieve and Refine: Improved Sequence Generation Models For Dialogue

https://towardsdatascience.com/text-classification-with-hugging-face-transformers-in-tensorflow-2-without-tears-ee50e4f3e7ed#:~:text=The%20Hugging%20Face%20transformers%20package,2%20is%20supported%20as%20well.

Text Classification with Hugging Face Transformers in TensorFlow 2 (Without Tears)

https://towardsdatascience.com/introduction-to-logistic-regression-66248243c148


Introduction to Logistic Regression

https://towardsdatascience.com/introduction-to-logistic-regression-66248243c148

https://towardsdatascience.com/support-vector-machines-a-brief-overview-37e018ae310f



Learn Before

Related

Learn After