Category: Data Science

Data Science, Machine Learning, Deep Learning, Data Scientist

How to Become a Good Agricultural Engineer

Agricultural engineering is a combination of engineering technology and biological science applied to the field of agriculture. World population will exceed 9.6 billion by 2050 and we need to increase agricultural production by 70 percent to meet demand.

Agricultural Engineers will be one of the main problem solvers.  They will be a bridge between technology and agriculture and it will be one leg of the solution.

Get Code To:

  • Get Agriculture Engineer job posts summaries from indeed.com
  • Save all summaries in one python list
  • Create WordCloud based on this list

Libraries Used:

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sat Oct 18 02:43:37 2017

@author: erayonler
"""
#Importing libraries
import requests
import matplotlib.pyplot as plt
from bs4 import BeautifulSoup
from wordcloud import WordCloud, STOPWORDS

#Scraping text data from indeed.com

text = []

for index in range(0,1000,10):
page = "https://www.indeed.com/jobs?q=agricultural+engineer&start="+str(index)

web_result = requests.get(page).text

soup = BeautifulSoup(web_result)

for listing in soup.findAll("span", {"class": "summary"}):
text.append(listing.text)

texts = " ".join(text)

stopwords = set(STOPWORDS)

#Creating and visualizing word cloud
wordcloud = WordCloud(width = 1000, height = 500, stopwords = stopwords).generate(texts)
plt.figure(figsize=(15,8))
plt.imshow(wordcloud)
plt.axis("off")
plt.show()
Agricultural Engineer Word Cloud

This word cloud can give us an intuition about what, companies demand from agricultural engineers.

For example from the first view, I notice that System Engineer, Test Engineer, Electrical Controls, Project Engineer, Agricultural Production, Software Engineer, Design Engineer have a higher frequency of agricultural engineer job posts.

Why don’t you copy/modify the code and try for different job search keywords?

How Algorithms Changing Our Lives?

https://www.kdnuggets.com/2016/09/great-algorithm-tutorial-roundup.html

Even we notice or not the algorithms are everywhere in our daily life. When we are making payment with credit card, while we are searching something at the search engine or while our smartphone camera automatically finds the faces at the camera and focusing on them, The algorithms are working in the background and making our lives easier.

We tend to think algorithms came into our lives by the computer age but it is wrong. Algorithms are helping us from the ancient times. For example, Euclid’s algorithm to find the greatest common divisor, which we learn at primary school math classes, probably one the oldest algorithm.

Algorithms are for solving the problem in structural, efficient and fastest way. So let’s look at some algorithms which make our lives easier.

PageRank Algorithm

One of the well-known algorithm of modern time is PageRank algorithm. It is a ranking algorithm and when you enter a query, it is sorting search results that you most likely interested in, based on ranking scores.

Sorting Algorithms

As it can be understood from the title, the job of the sorting algorithms are sorting things. Bubble sort algorithm is one the first iconic algorithm of all the time which is created in 1963 at System Development Corporation (First Software company of the world)

There are a lot of sorting algorithms for different purposes. For example merge sort, quick sort, insertion sort, selection sort and list goes on…

Sorting algorithms bring order to the world

“The Secret Rules of Modern Living: Algorithms”

Matching Algorithms

Online dating is so popular at these days. Even studies show that over third of marriages started online. These dating websites are using matching algorithms. They are searching through your profiles and match people up based on their likes, dislikes, and personality. Another research also shows couples who met online tend to be happier and have longer marriages.

In 2012 for the first time, Nobel Prize was awarded because of an algorithm  which is created by David Gale and  Lloyd Shapley.

This matching algorithm is also being used for matching organ donors, placing students in high schools and universities.

Shortest Path Algorithm

Sometimes problems are very complex to solve. For example very well known, traveling salesman problem.

The salesman must visit all cities in the shortest way. But how to do it? The first solution comes to our mind is brute force solution of this problem – trying all possible routes and then decide which one is the shortest.

But unfortunately when the number of cities increased possible routes are also increasing exponentially. When we look at the numbers we will see better:

  • 3 cities: There are only 3 possible routes
  • 5 cities: There are 60 possible routes
  • 6 cities: There are 360 possible routes
  • 10 cities: There are 1.8 million possible routes

https://www.hackerearth.com/practice/algorithms/graphs/shortest-path-algorithms/tutorial/

Machine Learning Algorithms

Now, we came into another breakpoint at technology. By using past data and computational power,  computers can learn like us. We don’t need to program them explicitly. If we feed them with enough and appropriate data; they can find patterns in and they can learn from data.

Programs don’t have frozen mind anymore.  They can learn and adapt themselves. We must be ready for the more automated world. Algorithms will shape our lives more and more in the future.

 

The Secret Rules of Modern Living: Algorithms.  I liked this documentary on Netflix and I strongly suggest you watch If you have free time and one cup of coffee.

How We Will Create Our Data Science Savvy Generation?

It is ok Data Science is the sexiest job of 21st century (Harvard Business Reviews) but our generation is really to ready to be data science savvy?

Source: HBR

Yes, it is very easy to see popularity of data science increasing like a rocket

Data Science Trend over years

But just this graph is not only explaining the whole story. Because when we look at top 5 related queries we saw a different face of the story.

Data Science Top Five Queries

It is very obvious there is a strong interest on data science but as we see on the queries “what is data science” is at 2nd and “what is data” is in 4th place of the query list. Actually this is really great thing because data science is very important for everyone while the world is becoming more digital (E.g. agriculture)

Our digital business research with MIT showed that 70% of CEOs believe their core business model is under attack, and 90% of them believe they do not yet have the right leadership team or technical skills to adapt.

If I could awake a curiosity about data science you can also check this great article by Russ Thompson. I think you will enjoy to read. I especially like this very simple but crystal clear definition

A data scientist is the adult version of the kid who can’t stop asking “Why?”.

https://blog.alexa.com/know-data-science-important/

The question is how will we make this transformation? How we will educate people about data science? By classical education? By online course? By bootcamps? or maybe digital revolution will lead us to more effective hybrid solutions?

How we will create our data science savvy generation?

Eray Onler

 

Machine Learning Can Help You To Find Plant Diseases

Smallholder family farming is the backbone of agriculture at developing countries. There are 500 millions smallholder farmer all around the world.  Pests and plant diseases are one of the biggest problem of farmers and these smallholders players usually don’t have enough chance to have consultancy. Luckily machine learning can help them about this. The app Plantix can be answer of this question. Just a simple smartphone can help you detect disease by taking photo of the plant. App. is not only detecting but also suggesting you possible solutions. Every new image is increasing recognition performance.

Plantix - AI-driven plant disease detection

Farmers Can Turn Their Data to Revenue Stream?

Data can be really revenue stream for farmers? Unless there is still a discussion about ownership about data

Who Owns Farmers’ Big Data?

https://www.forbes.com/sites/emc/2014/07/08/who-owns-farmers-big-data/#2263038b6033

Earning money from farm data is possible now. Technology startup FarMobile created DataStore to sell electronic farm records data. And they are harvesting farm data $1/acre.

If you would like to read more about this topic I suggest you to visit this article.

Definitely We Need Smarter Agriculture

We need to find a new and smarter ways for sustainable farming, while population is growing (it will be 9.8billions in 2050) and we are loosing fertile farmlands for urbanisation.

Crowd at the Palio horse race in Siena

Victor Spinelli | Photographer’s Choice | Getty
Crowd at the Palio horse race in Siena

This FAO report shows us importance of common vision for sustainable farming.

IoT is a way of connecting objects to internet. We can use this power for collecting more data from the farm and taking better and smarter decision.

I really like this article

The Internet of Grapes: How IoT and weather come together to support the global food system

thanks to its writer Karen Lewis

Farklı Kültür Bitkilerinin Renk Özelliklerinin Belirlenmesi Üzerine Bir Araştırma

Tarımda kimyasal mücadele en etkili uygulama tekniği olması yanında çevre ve insan sağlığı açısından olumsuz sonuçlarda yaratmaktadır. Klasik uygulamalarda tüm tarlaya ilaçlama yapılırken gereğinden fazla kimyasal tüketimi ortaya çıkmaktadır. Birçok araştırmacı buna yönelik birçok araştırmalar yapmıştır ve yapmaktadır. Bu çalışmada da fazla kimyasal tüketimini ortadan kaldırmaya yönelik sistemlerin geliştirilmesi için bazı özelliklerin belirlenmesi sağlanmıştır. Bu sistemin yapılabilmesi için, tarlada yabancı ot ve kültür bitkisinin ayırtedilebilmesini sağlayacak renk özelliklerinin bilinmesi gereklidir. Bu çalışmada bitki materyali olarak seçilen 3 farklı tip kültür bitkisi (ayçiçeği, domates, hıyar) yapraklarının renk özellikleri kamera ve renk ölçüm cihazı kullanılarak RGB renk uzayında incelenmiştir. Söz konusu kültür bitkilerini Ayçiçeği (Helianthus annuus L. L.), Domates (Solanum lycopersicum L.), Hıyar (Cucumis sativus L.) temsil etmektedir. Bitki yapraklarının kamera ile alınan görüntülerin renksel özellikleri, görüntü işleme tekniği kullanılarak incelenmiştir. Renk ölçüm cihazı ile elde edilen renksel özellikler ise kamera ile elde edilen değerlerin karşılaştırılması amacıyla kullanılmıştır. Tüm sonuçlar incelendiğinde, kamera ve renk ölçüm cihazında ki r, g ve b değerleri (aydınlıktaki değişimin ve yaprak üzerine düşen gölgeleri gidermek amacıyla elde edilen değerler) karşılaştırıldığın da ayçiçeği bitkisinin, r, g ve b değerleri arasında önemli derecede farklılık olduğu, domates ve hıyar bitkilerinde ise önemli derecede ki fark b değerlerinde olduğu görülmüştür.

You can download full article from tarmek

 

Kendi Yürür Bir Prototip Tarımsal Araç Tasarımı

Bu proje kapsamında özellikle sıraya ekilen tarımsal ürünlerde kullanılabilecek bir kendi yürür tarımsal araç prototipi geliştirilmiştir. Bu araç otomatik kontrol içeren ve tekerlek kullanıma alternatif bacakları olan, her yöne hareket edebilen bir güç kaynağıdır. Sistemi yönlendiren yazılım yardımıyla farklı komutlar yüklenerek farklı fonksiyonlara sahip olabilmektedir. Bu fonksiyonlar çekme, taşıma, bitki tanıma, var/yok kararı, görüntü işleme, ses algılama gibi özel amaçlara göre geliştirilebilir. Temel olarak arasında gidebilmekte, sıra bittiğinde kendi kararı ile dönebilmekte ve istediği sıraya girebilmektedir. Projede geliştirilen araç bir bataryadan aldığı enerji ile çalışmaktadır. Robot ana bilgisayar, kamera, laser tarayıcı sensör, servo motorlar ve arduino mega temel elemanlarından oluşmaktadır. Yazılımın yer aldığı ana bilgisayar aracılığıyla Laser tarayıcı ve kameradan gelen verilerin tüm hesaplama ve karar verme işlemi yapmaktadır. Ayrıca ses komutu ile bazı yönlendirmeler yapılabilmektedir.. Araç tekerlek yerine bacaklar üzerinde hareket etmekte ve ilerlemektedir. Robotun hareketi bacaklarda kullanılan servo motorların belli sıra ile çalıştırılması ile sağlanmaktadır. Merkezi programda belirlenen hareket şekli servo motorlara arduino mega mikro-denetleyici ünite ile aktarılmaktadır. Robot bir operatör yardımıyla da kullanılabilmektedir. Ek olarak karşılaşılabilecek istenmeyen durumlarda komuta edebilmek için Operatör robotu bir joystick ile veya sesli komutlar ile uzaktan kontrol edilebilmektedir.

Self propelled agricultural vehicle prototype is developed for row-crop cultivation at this project. This robot is moving automaticly with legs . System can be programmed for different purposes and functions. These functions are transport, plant identification, weed detection. Basically, the robot, can go between rows, and can turn to next row at the end of the rows. The robot is working with the energy from the battery. The robot consists host computer, camera, laser scanner sensor, servo motors and arduino mega.The software at the host computer receive data from laser scanner and camera, makes all the calculations and decision-making process. In addition, some guidance can be done with voice commands . The robot is moving with legs instead of wheels. Walking of the robot is provided by sequential operations servo motors used in the legs. Servo motors are driving by arduino mega microprocessor and driver circuit. The robot can also be used with the help of an operator. In addition to the operator to be able to command in undesirable situations, robot can be controlled remotely with a joystick or voice commands.

http://acikerisim.nku.edu.tr:8080/xmlui/handle/20.500.11776/2109

Determination of energy balance of apple (Malus Domestica) production in Turkey: A case study for Tekirdag province

This study aims to define the energy usage efficiency in apple cultivation in the Province of Tekirdağ. The study was conducted during 2015 production season through observation and measurement in an apple garden with a land area of 12 da and located in Nusratlı village in Central Tekirdağ. It has been tried to reveal the role of mechanization energy among all the inputs. According to the calculated data, in apple cultivation the respective figures for total energy input, total fruition, total energy output, energy output/input rate, specific energy, energy productivity and net energy have been calculated as 58839.65 MJ ha-1, 38370 kg ha-1, 92088.00  MJ ha-1, 1.56, 1.53 MJ kg-1, 0.65 kg MJ-1 and 33248.35 MJ ha-1 respectively. As a result, among the general energy inputs in apple cultivation, the highest energy consuming items have been respectively defined as fertilizer energy, fuel-oil energy, chemicals, machinery, human labour and irrigation energy.

 

http://dergipark.gov.tr/omuanajas/issue/27219/289604

Crowdflower – AI on your business

There is no doubt data science is really very trend topic at last years and it is completely changing our life. Many experts say data is new oil and it is not really an abstract.

Source: https://medium.com/towards-data-science/data-is-the-new-oil-f11440e80dd0

But most of the existing data is raw and most of the time we need an human in ai-loop. Last week I saw a company who is working about this and I especially appreciate A Guide to Machine Learning for Computer Vision Algorithms. I think you can also follow  and take a look at use cases of this company CrowdFlower