Large-scale cancer drug sensitivity data have become available for a collection of cancer cell lines, but only limited drug response data from patients are available. Bridging the gap in pharmacogenomics knowledge bet...
详细信息
A growing number of machine learning (ML) projects in manufacturing require the collaboration of various experts. In addition to data scientists, stakeholders with production engineering knowledge have to specify and ...
详细信息
A growing number of machine learning (ML) projects in manufacturing require the collaboration of various experts. In addition to data scientists, stakeholders with production engineering knowledge have to specify and prioritize individual project tasks. data engineers prepare input data, while machine learning operations (MLOps) engineers ensure that trained models are deployed and monitored within IT landscapes. Existing project management approaches, e.g., Scrum, have problems for ML projects, as they do not consider various expert roles or ML project stages. We propose a project management approach defining a Kanban workflow by readjusting stages of ML development lifecycles, e.g., CRISP DM. This makes it possible to map expert roles to stages of the Kanban workflow. An adapted Kanban board allows visualizing and reviewing the status of all project tasks. We validate our approach with specific use cases, showing that it facilitates ML project management in manufacturing.
This study introduces the Quantum-Train Quantum Fast Weight programmer (QT-QFWP) framework, enabling efficient and scalable programming of variational quantum circuits (VQCs) through quantum-driven parameter updates f...
详细信息
With the popularity of the current Internet age, online social platforms have provided a bridge for communication between private companies, public organizations, and the public. The purpose of this research is to und...
详细信息
With an increasing interest in the digitization effort of ancient manuscripts, ancient character recognition becomes one of the most important areas in the automated document image analysis. In this regard, we propose...
详细信息
With an increasing interest in the digitization effort of ancient manuscripts, ancient character recognition becomes one of the most important areas in the automated document image analysis. In this regard, we propose a Convolutional Neural Network (CNN)-based classifier to recognize the ancient Sundanese characters obtained from a digital collection of Southeast Asian palm leaf manuscripts. In this work, we utilize two different preprocessing techniques for the dataset. The first technique involves the use of geometric transformations, noise background addition, and brightness adjustment to augment the imbalanced samples to be fed into the classifier. The second technique makes use of the Otsu’s threshold method to binarize the characters and only uses the usual geometric transformations for the data augmentation. The proposed network with different data augmentation processes is trained on the training set and tested on the testing set. Image binarization from the second technique can outperform the performance of the CNN-based classifier upon the first technique by achieving a testing accuracy of 97.74%.
Adopting a deep learning model into bird sound classification tasks becomes a common practice in order to construct a robust automated bird sound detection system. In this paper, we employ a four-layer Convolutional N...
详细信息
Adopting a deep learning model into bird sound classification tasks becomes a common practice in order to construct a robust automated bird sound detection system. In this paper, we employ a four-layer Convolutional Neural Network (CNN) formulated to classify different species of Indonesia scops owls based on their vocal sounds. Two widely used representations of an acoustic signal: log-scaled mel-spectrogram and Mel Frequency Cepstral Coefficient (MFCC) are extracted from each sound file and fed into the network separately to compare the model performance with different inputs. A more complex CNN that can simultaneously process the two extracted acoustic representations is proposed to provide a direct comparison with the baseline model. The dual-input network is the well-performing model in our experiment that achieves 97.55% Mean Average Precision (MAP). Meanwhile, the baseline model achieves a MAP score of 94.36% for the mel-spectrogram input and 96.08% for the MFCC input.
The number of findings in cancer genomics research has grown rapidly in the last decade due to the decline in the cost of human sequencing and genotyping. However, the majority of the reported significant marker assoc...
详细信息
The number of findings in cancer genomics research has grown rapidly in the last decade due to the decline in the cost of human sequencing and genotyping. However, the majority of the reported significant marker associated with cancer traits are based on European and East Asian population. Large population such as South Asian and South-East Asian population are under-represented in genomics research. In this study, we explored the possibility of computing a Polygenic Risk Score (PRS) of colorectal cancer on our test sample based on reported significant Single Nucleotide Polymorphism (SNP). The SNPs used to compute the risk score were collected from GWAS Central and GWAS Catalog. Significant SNPs from IC3 study were used as a benchmark. The result shows that calculating colorectal cancer risk score using reported significant marker from different population group is possible. The p-value of our statistic model shows significant differences between case and control group risk score.
This study aims to address the common issue of biased estimation errors in time series modeling by analyzing the error in locating ideal hyperparameters and defining appropriate validation methods. Specifically, it fo...
详细信息
In recent years, research in human counting from CCTV (Closed Circuit Television) images have found an increasing demand to be deployed in real-world applications. The applications have been implemented in various set...
详细信息
In recent years, research in human counting from CCTV (Closed Circuit Television) images have found an increasing demand to be deployed in real-world applications. The applications have been implemented in various settings, both indoor and outdoor. In the case of indoor setting, we found a type of room setting that conveys a problem to human counting model if we need to count only humans inside a room. With this respect, we present RHC (Room Human Counting) dataset, which images are captured in the aforementioned setting. The dataset can be used to develop a robust model that can differentiate between humans inside and outside a room. The dataset is publicly available at https://***/datasets/vt5c8h6kmh/1.
To strengthen conservation efforts for preserving biodiversity in a conservation area, forest inventory is important to understand the natural succession process in the area and to establish a monitoring strategy. Fur...
详细信息
ISBN:
(纸本)9781665453967
To strengthen conservation efforts for preserving biodiversity in a conservation area, forest inventory is important to understand the natural succession process in the area and to establish a monitoring strategy. Further, tree inventory aims to monitor the output yielded in the area. More specifically, the tree inventory in the watershed area plays a key role to achieve Sustainable Development Goals (SDG), especially in riparian zones which are also vital parts of green zones in forests. However, the traditional inventory approach is time-consuming and laborious therefore the development of an expert system to assist in inventory monitoring is required. In this study, we develop a monitoring system via a mobile application to collect, analyze and visualize tree inventory data. The application includes algorithms required to compute tree biodiversity, distribution, and richness for the given input of the data of all tree species in a conservation area. For the model validation stage, we compare the traditional inventory approach with our proposed application-based approach to compute diversity inventory in two riparian locations: Klaten Conservation Park and Wonosobo Conservation Park. After the three-day data collection in the areas, we obtain that the accuracy of reading data of our proposed system can achieve more than 90% in comparison with the manual approach. This demonstrates that the system can assist forestry workers to perform more efficient tree inventories in different locations.
暂无评论