David Camilo Corrales

An Empirical Multi-classifier for Coffee Rust Detection in Colombian Crops

Rust is a disease that leads to considerable losses in the worldwide coffee industry. In Colombia, the disease was first reported in 1983 in the department of Caldas. Since then, it spread rapidly through all other coffee departments in the country. Recent research efforts focus on detection of disease incidence using computer science techniques such as supervised learning algorithms. However, a number of different authors demonstrate that results are not sufficiently accurate using a single classifier. Authors in the computer field propose alternatives for this problem, making use of techniques that combine classifier results. Nevertheless, the traditional approaches have a limited performance due to dataset absence. Therefore we proposed an empirical multi-classifier for coffee rust detection in Colombian crops.

How to Address the Data Quality Issues in Regression Models: A Guided Process for Data Cleaning

Today, data availability has gone from scarce to superabundant. Technologies like IoT, trends in social media and the capabilities of smart-phones are producing and digitizing lots of data that was previously unavailable. This massive increase of data creates opportunities to gain new business models, but also demands new techniques and methods of data quality in knowledge discovery, especially when the data comes from different sources (e.g., sensors, social networks, cameras, etc.). The data quality process of the data set proposes conclusions about the information they contain. This is increasingly done with the aid of data cleaning approaches. Therefore, guaranteeing a high data quality is considered as the primary goal of the data scientist. In this paper, we propose a process for data cleaning in regression models (DC-RM). The proposed data cleaning process is evaluated through a real datasets coming from the UCI Repository of Machine Learning Databases. With the aim of assessing the data cleaning process, the dataset that is cleaned by DC-RM was used to train the same regression models proposed by the authors of UCI datasets. The results achieved by the trained models with the dataset produced by DC-RM are better than or equal to that presented by the datasets’ authors.

Lack of Data: Is It Enough Estimating the Coffee Rust with Meteorological Time Series?

Rust is the most economically important coffee disease in the world. Coffee rust epidemics have affected a number of countries including: Colombia, Brazil and Central America. Researchers try to predict the Incidence Rate of Rust (IRR) through supervised learning models, nevertheless the available IRR measurements are few, then the data set does not represent a sample trustworthy of the population. In this paper we use Cubic Spline Interpolation algorithm to increase the measurements of Incidence Rate of Rust and subsequently we construct different subsets of meteorological time series: (i) Daily meteorology, (ii) Meteorological variation, and (iii) Previous meteorology using M5 Regression Tree, Support Vector Regression and Multi-Layer Perceptron. Previous meteorology with Multi-Layer Perceptron have shown better results in measures as Pearson Coefficient Correlation of 0.81 and Mean Absolute Error \(=\) 7.41%.

Validation of Coffee Rust Warnings Based on Complex Event Processing

The rust is the main coffee crop disease in the world. In the Colombian and Brazilian plantations, the damage leads to a yield reduction of 30 % and 35 % respectively in regions where the meteorological conditions are propitious to the disease. Recently, researchers have focused on detecting the coffee rust disease starting from climate monitoring and parameters of crop control; however most of the monitoring systems lack the ability to process multiple source information and analyse it in order to identify abnormal situations and validate the generated warnings. In this paper, we propose a CEP engine and a prediction system integration for early warning systems applied to the coffee rust detection, capable of analysing multiple incoming events from the monitoring system and validating the warnings detection; evaluating an experimental prototype in a field test with satisfactory results.

A Cloud-Based Platform for Decision Making Support in Colombian Agriculture: A Study Case in Coffee Rust

In the last years, the yield of Colombian crops has been affected by climate change. The weather variation affects the Colombian crops with the occurrence of diseases as coffee rust. To address the coffee rust control, we proposed a cloud-based platform for decision making support named AgroCloud. The coffee crop weather of 100 municipalities from upper basin of the Cauca river were monitored. This information was used to improve the disease control process. User Acceptance Test carried out with domain end users show that the platform is useful and is easily usable.

Water quality detection based on a data mining process on the California estuary

Freshwater is considered one of the most important renewable natural resources of the planet. In this sense, it is vital to study and evaluate the water quality in rivers and basins. The USA and especially the border states like California face the same water problems as its southern neighbours, such as the deterioration of public drinking water systems and the continued appearance of pollutants that threaten domestic water sources. This implies the need to monitor and analyse the water supplies in each region. Several researches have been conducted to develop water quality detection systems through supervised learning algorithms. However, these research approaches set aside the data processing to improve the performance of supervised learning algorithms. This paper presents an improvement of data processing techniques for a water quality detection system based on supervised learning and data quality techniques for the California estuary.

A Guideline for Building Large Coffee Rust Samples Applying Machine Learning Methods

Coffee rust has become a serious concern for many coffee farmers and manufacturers. The American Phytopathological Society discusses its importance saying this: “the most economically important coffee disease in the world,” while “in monetary value, coffee is the most important agricultural product in international trade”. The early detection has inspired researchers to apply supervised learning algorithms on predicting the disease appearance. However, the main drawback of the related works is the few data samples of the dependent variable: Incidence Rate of Rust, since the datasets do not have a reliable representation of the disease, which will generate inaccurate classifiers. This paper provides a guide to increase coffee rust samples applying machine learning methods through a systematic review about coffee rust in order to select appropriate algorithms to increase rust samples.

Data Processing for a Water Quality Detection System on Colombian Rio Piedras Basin

Freshwater is considered one of the most important of planet’s renewable natural resources. In this sense, it is vital to study and evaluate the water quality in rivers and basins. A study area is Rio Piedras Basin, which is the main water supplier source of 9 rural communities in Colombia. Nevertheless, these communities do not make a water quality control. Different research has been conducted to develop water quality detection systems through supervised learning algorithms. However, these research approaches set aside the data processing for improve the outcomes of supervised learning algorithms. This paper presents an improvement of data processing techniques for a water quality detection system based on supervised learning and data quality techniques for Rio Piedras Basin.