Mediaboss Marketing

The expectation is that in 2021, artificial intelligence and machine learning technologies will continue to become more mainstream. Businesses that havent traditionally viewed themselves as candidates for AI applications will embrace these technologies.

A great story of machine learning being used in an industry that is not known for its technology investments is the story of Makoto Koike. Using Googles TensorFlow, Makoto initially developed a cucumber sorting system using pictures that he took of the cucumbers. With that small step, a machine learning cucumber sorting system was born.

Getting started with AI and machine learning is becoming increasingly accessible for organizations of all sizes. Technology-as-a-service companies including Microsoft, AWS and Google all have offerings that will get most organizations started on their AI and machine learning journeys. These technologies can be used to automate and streamline manual business processes that have historically been resource-intensive.

An article on forbes.com claims that, as business leaders continue to refine their processes to support the new normal of the Covid-19 pandemic, they should be considering where these technologies might help reduce manual, resource-intensive or paper-based processes. Any manual process should be fair game for review for automation possibilities.

Photo source: Dreamstime.com

Read the original:
Using AI and Machine Learning will increase in horti industry - hortidaily.com

Researchers have developed a machine learning-based approach to identify drugs that might be repurposed to fight COVID-19 in elderly patients.

The machine learning-based approach developed by researchers at the Massachusetts Institute of Technology (MIT; Cambridge, MA, USA) aims to identify drugs already on the market that could potentially be repurposed to fight COVID-19, particularly in the elderly. The system accounts for changes in gene expression in lung cells caused by both the disease and aging. That combination could allow medical experts to more quickly seek drugs for clinical testing in elderly patients, who tend to experience more severe symptoms. The researchers pinpointed the protein RIPK1 as a promising target for COVID-19 drugs, and they identified three approved drugs that act on the expression of RIPK1.

Stiffening lung tissue in COVID-19 harmed older patients due to aging shows different patterns of gene expression than in younger people, even in response to the same signal. The researchers looked at aging together with SARS-CoV-2, including identifying the genes at the intersection of these two pathways. To select approved drugs that might act on these pathways, the team turned to big data and artificial intelligence.

The researchers zeroed in on the most promising drug repurposing candidates in three broad steps. First, they generated a large list of possible drugs using a machine-learning technique called an autoencoder. Next, they mapped the network of genes and proteins involved in both aging and SARS-CoV-2 infection. Finally, they used statistical algorithms to understand causality in that network, allowing them to pinpoint "upstream" genes that caused cascading effects throughout the network. In principle, drugs targeting those upstream genes and proteins should be promising candidates for clinical trials.

To generate an initial list of potential drugs, the team's autoencoder relied on two key datasets of gene expression patterns. One dataset showed how expression in various cell types responded to a range of drugs already on the market, and the other showed how expression responded to infection with SARS-CoV-2. The autoencoder scoured the datasets to highlight drugs whose impacts on gene expression appeared to counteract the effects of SARS-CoV-2. Next, the researchers narrowed the list of potential drugs by homing in on key genetic pathways. They mapped the interactions of proteins involved in the aging and Sars-CoV-2 infection pathways. Then they identified areas of overlap among the two maps. That effort pinpointed the precise gene expression network that a drug would need to target to combat COVID-19 in elderly patients.

The researchers were yet to identify which genes and proteins were "upstream" (i.e. they have cascading effects on the expression of other genes) and which were "downstream" (i.e. their expression is altered by prior changes in the network). An ideal drug candidate would target the genes at the upstream end of the network to minimize the impacts of infection. So the team used algorithms that infer causality in interacting systems to turn their undirected network into a causal network. The final causal network identified RIPK1 as a target gene/protein for potential COVID-19 drugs, since it has numerous downstream effects. The researchers identified a list of the approved drugs that act on RIPK1 and may have potential to treat COVID-19. Previously these drugs have been approved for the use in cancer. Other drugs that were also identified, including ribavirin and quinapril, are already in clinical trials for COVID-19.

The researchers now plan to share their findings with pharmaceutical companies, clinical testing is needed to determine efficacy before any of the identified drugs can be approved for repurposed use in elderly COVID-19 patients,. While this particular study focused on COVID-19, the researchers say their framework is extendable.

"I'm really excited that this platform can be more generally applied to other infections or diseases," said Anastasiya Belyaeva, study co-author and MIT PhD student.

Related Links:Massachusetts Institute of Technology (MIT)

See original here:
New Machine Learning-Based Approach Identifies Existing Drugs That Could Be Repurposed to Fight COVID-19 - HospiMedica

IoT-enabled Agricultural (IoTAg) monitoring is smart, connected agriculture's fastest-growing ... [+] technology segment projected to reach $4.5 billion by 2025, according to PwC.

AI, machine learning (ML) and the IoT sensors that provide real-time data for algorithms increase agricultural efficiencies, improve crop yields and reduce food production costs. According to the United Nations' prediction data on population and hunger, the world's population will increase by 2 billion people by 2050, requiring a 60% increase in food productivity to feed them. In the U.S. alone, growing, processing and distributing food is a $1.7 trillion business, according to the U.S. Department of Agriculture's Economic Research Service. AI and ML are already showing the potential to help close the gap in anticipated food needs for an additional 2 billion people worldwide by 2050.

Agriculture Is One Of The Most Fertile Industries There Are For AI & Machine Learning

Imagine having at least 40 essential processes to keep track of, excel at and monitor at the same time across a large farming area often measured in the hundreds of acres. Gaining insight into how weather, seasonal sunlight, migratory patterns of animals, birds, insects, use of specialized fertilizers, insecticides by crop, planting cycles and irrigation cycles all affect yield is a perfect problem for machine learning. How financially successful a crop cycle has never been more dependent on excellent data. That's why farmers, co-ops and agricultural development companies are doubling down on data-centric approaches and expanding the scope and scale of how they use AI and machine learning to improve agricultural yields and quality. The following are ten ways AI has the potential to improve agriculture in 2021:

1.Using AI and machine learning-based surveillance systems to monitor every crop field's real-time video feeds identifies animal or human breaches, sending an alert immediately.AI and machine learning reduce domestic and wild animals' potential to accidentally destroy crops or experience a break-in or burglary at a remote farm location. Given the rapid advances in video analytics fueled by AI and machine learning algorithms, everyone involved in farming can protect their fields and buildings' perimeters. AI and machine learning video surveillance systems scale just as easily for a large-scale agricultural operation as for an individual farm.Machine-learning based surveillance systems can be programmed or trained over time to identify employees versus vehicles. Twenty20 Solutions is a leader in the field of AI and machine learning-based surveillance and has proven effective in securing remote facilities, optimizing crops and deterring trespassers by using machine learning to identify employees who work onsite. An example of Twenty20 Solutions' real-time monitoring is shown here:

Relying on AI and machine learning algorithms to identify people and vehicles is streamlining remote ... [+] operations for agricultural businesses globally today.

2.AI and machine learning improve crop yield prediction through real-time sensor data and visual analytics data from drones. The amount of data being captured by smart sensors and drones providing real-time video streaming provides agricultural experts with entirely new data sets they've never had access to before. It's now possible to combine in-ground sensor data of moisture, fertilizer and natural nutrient levels to analyze growth patterns for each crop over time. Machine learning is the perfect technology to combine massive data sets and provide constraint-based advice for optimizing crop yields. The following is an example of how AI, machine learning, in-ground sensors, infrared imagery and real-time video analytics all combine to provide farmers with new insights into how they can improve crop health and yields:

Drones are proving to be a reliable platform for capturing data on how specific fertilizers, ... [+] watering patterns and pesticide treatment methods are improving crop yields.

3.Yield mapping is an agricultural technique that relies on supervised machine learning algorithms to find patterns in large-scale data sets and understand the orthogonality of them in real-time all of which is invaluable for crop planning. Its possible to know the potential yield rates of a given field before a vegetation cycle is ever started. Using a combination of machine learning techniques to analyze 3D mapping, social condition data from sensors and drone-based data of soil color, agricultural specialists can now predict the potential soil yields for a given crop. A series of flights are completed to get the most accurate data set possible. The following graphic shows the result of a yield mapping analysis:

Supervised and unsupervised machine learning algorithms are being used to determine how best to ... [+] optimize yields by field.

4.The UN, international agencies and large-scale agricultural operations are pioneering drone data combined with in-ground sensors to improve pest management. Using infrared camera data from drones combined with sensors on fields that can monitor plants' relative health levels, agricultural teams using AI can predict and identify pest infestations before they occur. An example of this is how the UN is using working in conjunction with PwC to evaluate data palm orchards in Asia for potential pest infestations, as is shown in the image below:

The UN is combining on-ground sensor and drone data to fine-tune their machine learning algorithms ... [+] that assist farmers in achieving greater yields from the crops.

5.Today, theres a shortage of agricultural workers, making AI and machine learning-based smart tractors, agribots and robotics a viable option for many remote agricultural operations that struggle to find workers.Large-scale agricultural businesses cant find enough employees and turn to robotics for hundreds of acres of crops while also providing an element of security around the perimeter of remote locations. Programming self-propelled robotics machinery to distribute fertilizer on each row of crops helps keep operating costs down and further improve field yields. Agriculture robots sophistication has grown quickly, an example of which is shown in the dashboard of the VineScout robot in use.

Agricultural robotics are proving to be adept at capturing valuable data for fine-tuning AI and ... [+] machine learning algorithms, further improving crop yields.

6.Improving the track-and-traceability of agricultural supply chains by removing roadblocks to getting fresher, safer crops to market is a must-have today. The pandemic accelerated track-and-traceability adoption across all agricultural supply chains in 2020 and will continue to drive its adoption this year. A well-managed track-and-trace system helps reduce inventory shrinkage by providing greater visibility and control across supply chains. A state-of-the-art track-and-trace system can differentiate between inbound shipments' batch, lot and container level assignments of materials. Most advanced track-and-trace systems rely on advanced sensors to gain greater knowledge of each shipment's condition. RFID and IoT sensors are now becoming more commonplace across manufacturing. Walmart ran a pilot to see how RFID could streamline a distribution center's track-and-trace performance and improved efficiency by 16 times over manual methods.

7.Optimize the right mix of biodegradable pesticides and limiting their application to only the field areas that need treatment to reduce costs while increasing yields is one of the most common uses of AI and machine learning in agriculture today. By using intelligent sensors combined with visual data streams from drones, agricultural AI applications can now detect a planting area's most infected areas. Using supervised machine learning algorithms, they can then define the optimal mix of pesticides to reduce pests' threat spreading further and infecting healthy crops.

8.Price forecasting for crops based on yield rates that help predict total volumes produced are invaluable in defining pricing strategies for a given crop. Understanding yield rates and quality levels of crops help agricultural firms, co-ops and farmers better negotiate for the best possible price for their harvests. Considering the total demand for a given crop to determine if the price elasticity curve for a given crop is inelastic, unitary, or highly elastic defines what the pricing strategy will be. Knowing this data alone saves agricultural businesses millions of dollars a year in lost revenue.

9.Finding irrigation leaks, optimizing irrigation systems and measuring how effective frequent crop irrigation improves yield rates are all areas AI contributes to improving farming efficiencies. Water is the scarcest resource in many parts of North America, especially in communities that rely most on agriculture as their core business. Being efficient in using it can mean the difference between a farm or agricultural operation staying profitable or not. Linear programming is often used to calculate the optimal amount of water a given field or crop will need to reach an acceptable yield level. Supervised machine learning algorithms are ideal for ensuring fields and crops get enough water to optimize yields without wasting any in the process.

10.Monitoring livestocks health, including vital signs, daily activity levels and food intake, ensures their health is one of the fastest-growing aspects of AI and machine learning in agriculture. Understanding how every type of livestock reacts to diet and boarding conditions is invaluable in understanding how they can be best treated for the long-term. Using AI and machine learning to understand what keeps daily cows contended and happy, producing more milk is essential. For many farms who rely on cows and livestock, this area opens up entirely new insights into how farms can be more profitable.

Excerpt from:
10 Ways AI Has The Potential To Improve Agriculture In 2021 - Forbes

Machine-learning techniques are required to improve the accuracy of predictive models. Depending on the nature of the business problem being addressed, there are different approaches based on the type and volume of the data. In this section, we discuss the categories of machine learning.

Supervised learning

Supervised learning typically begins with an established set of data and a certain understanding of how that data is classified. Supervised learning is intended to find patterns in data that can be applied to an analytics process. This data has labeled features that define the meaning of data. For example, you can create a machine-learning application that distinguishes between millions of animals, based onimages and written descriptions.

Unsupervised learning

Unsupervised learning is used when the problem requires a massive amount of unlabeled data. For example, social media applications, such as Twitter, Instagram and Snapchat, all have large amounts of unlabeled data. Understanding the meaning behind this data requires algorithms that classify the data based on the patterns or clusters it finds.

Unsupervised learning conducts an iterative process, analyzing data without human intervention. It is used with email spam-detecting technology. There are far too many variables in legitimate and spam emails for an analyst to tag unsolicited bulk email. Instead, machine-learning classifiers, based on clustering and association, are applied to identify unwanted email.

Reinforcement learning

Reinforcement learning is a behavioral learning model. The algorithm receives feedback from the data analysis, guiding the user to the best outcome. Reinforcement learning differs from other types of supervised learning, because the system isnt trained with the sample data set. Rather, the system learns through trial and error. Therefore, a sequence of successful decisions will result in the process being reinforced, because it best solves the problem at hand.

Deep learning

Deep learning is a specific method of machine learning that incorporates neural networks in successive layers to learn from data in an iterative manner. Deep learning is especially useful when youre trying to learn patterns from unstructured data.

Deep learning complex neural networks are designed to emulate how the human brain works, so computers can be trained to deal with poorly defined abstractions and problems. The average five-year-old child can easily recognize the difference between his teachers face and the face of the crossing guard. In contrast, the computer must do a lot of work to figure out who is who. Neural networks and deep learning are often used in image recognition, speech, and computer vision applications.

Continued here:
Machine Learning | IBM

For the past few years, JPMorgan has been busy building out its machine learning capability underDaryush Laqab, its San Francisco-based head of AI/machine learning products, who was hired from Google in 2019. Last time we looked, the bank seemed to be paying salaries of $160-$170k to new joiners onLaqab's team.

If that sounds appealing, you might want to watch the video below so that you know what you're getting into. Recorded at the AWS re:Invent conferencein December, it's only just made it to you YouTube. The video is flagged as a day in the life of JPMorgan's machine learning data scientists, butLaqab arguably does a better of job of highlighting some of the constraints data professionals at allbanks have to work under.

"There are some barriers to smooth data science at JPMorgan," he explains - a bank is not the same as a large technology firm.

For example, data scientists at JPMorgan have to check data is authorized for use, saysLaqab: "They need to go to a process to log that use and make surethat they have the adequate approvals for that intent in terms of use."

They also have to deal with the legacy infrastructureissue: "We are a large organization, we have a lot of legacy infrastructure," says Laqab. "Like any other legacy infrastructure, it is built over time,it is patched over time. These are tightly integrated,so moving part or all of that infrastructure to public cloud,replacing rule base engines with AI/ML based engines.All of that takes time and brings inertia to the innovation."

JPMorgan's size and complexity is another source of inertia as multiple business lines in multiple regulated entities in different regulated environments need to be considered. "Making sure that those regulatory obligationsare taken care of, again, slows down data science at times," saysLaqab.

And then there are more specific regulations such as those concerning model governance. At JPMorgan, a machine learning model can't go straight into a production environment."It needs to go through a model review and a model governance process," says Laqab. "- To make sure we have another set of eyes that looksat how that model was created, how that model was developed..." And then there are software governance issues too.

Despite all these hindrances, JPMorgan has already productionized AI models and built an 'Omni AI ecosystem' to help employees to identify and ingest minimum viable data so that they canbuild models faster. Laqab saysthe bank saved $150m in expenses in 2019 as a result. JPMorgan's AI researchers are now working on everything fromFAQ bots and chat bots, to NLP search models for the bank'sown content, pattern recognition in equities markets and email processing. - The breadth of work on offer is considerable. "We play in every market that is out there," saysLaqab,

The bank has also learned that the best way to structure its AI team is to split people into data scientists who train and create models and machine learning engineers who operationalize models, saysLaqab. - Before you apply, you might want to consider which you'd rather be.

Photo by NeONBRAND on Unsplash

Have a confidential story, tip, or comment youd like to share? Contact:sbutcher@efinancialcareers.comin the first instance. Whatsapp/Signal/Telegram also available. Bear with us if you leave a comment at the bottom of this article: all our comments are moderated by human beings. Sometimes these humans might be asleep, or away from their desks, so it may take a while for your comment to appear. Eventually it will unless its offensive or libelous (in which case it wont.)

Read more from the original source:
JPMorgan's head of machine learning explained what it's like to work there - eFinancialCareers