Identifying skeletal remains has been and will remain a challenge for forensic doctors and forens... more Identifying skeletal remains has been and will remain a challenge for forensic doctors and forensic anthropologists, especially in disasters with multiple victims or skeletal remains in an advanced stage of decomposition. This study proposes a machine learning method to determine gender starting from morphometric analysis of L1-L5 lumbar vertebrae in a modern Romanian population. The purpose of the present study was to observe whether by using the ML method there is a good predictability of gender in forensic identification based on parameters obtained from the metric analysis of the lumbar spine specific to the Romanian population. This paper offers two models of ML, RF and XGB, each with its own characteristics, and presenting different performance, random forest having the best. For both, we used two metrics (accuracy and roc_auc), the latter being the most used to highlight model performance. The L1-L5 lumbar vertebrae exhibit sexual dimorphism and can be used in gender estimati...
Introduction: Forensic identification practice requires a multitude of methods for positive ident... more Introduction: Forensic identification practice requires a multitude of methods for positive identification, among which the facial reconstruction occupies a high place. For this to be done in the best way possible, it is mandatory that each population have their own database regarding soft tissue thickness. Objective: The present research is focused on creating such a database for the Romanian adult population and on exploring the statistical distribution and correlation of craniometric landmarks, overall, and in groups defined according to sex and weight categories. This might provide support in estimating the values of missing landmarks or other variables (e.g., age, sex) for unidentified bodies. Methods: Twelve craniometric landmarks were measured on 100 cadavers less than 24 h after death. Results and discussions: Of the 12 anatomical landmarks, only one appears to manifest statistically significant differences between sexes, and seven landmarks appear to vary significantly amon...
Energy management systems from Romania do not have the capabilities of energy specific management... more Energy management systems from Romania do not have the capabilities of energy specific management due to lack of technology for real-time monitoring. As was the case in many other countries, the advent of smart metering technology will increase the level of energy data significantly. Therefore, the purpose of this paper is to present solutions that need to be taken to solve problems linked with the increasing amount of data recorded by sensors. For a better demonstration of theoretical elements exposed, we considered a data warehouse specific to utility companies. Section 2 of this article defines the three widely used parallel data warehouse architectures, while in Section 3 we clarify what architecture is suited to develop a data warehouse in the cloud. In the last part we transposed our tables in a "shared-nothing" architecture, trying to analyze queries performance.
Proceedings of the 19th International Conference on Computer Systems and Technologies, 2018
Big Data technologies brought the idea of parallel processing on cheaper commodity servers. When ... more Big Data technologies brought the idea of parallel processing on cheaper commodity servers. When dealing with huge amount of data, instead of migrating to more performant and costly hardware platforms, or buying resources in cloud, it is more affordable to add a number of cheaper servers as nodes for data processing and/or storage. NoSQL data stores, Hadoop ecosystems, NewSQL platforms have proved viable for Big Data storage and processing. In this paper we were concerned with setting up a platform for big data processing using commodity workstations. Many small and medium sized companies have limited resources and their workstations remain unused for more than 12 hours a day. Here Beowulf Cluster Computing could prove useful. Apache Impala was installed as part of a Hadoop distribution on a 9-node cluster. Three TPC-H database schema were loaded for the scale factors of 1, 2 and 10GB. A series of 100 SQL queries were randomly generated and executed for each scale factor. Results we...
Big Data systems manage and process huge volumes of data constantly generated by various technolo... more Big Data systems manage and process huge volumes of data constantly generated by various technologies in a myriad of formats. Big Data advocates (and preachers) have claimed that, relative to classical, relational/SQL Data Base Management Systems, Big Data technologies such as NoSQL, Hadoop and in-memory data stores perform better. This paper compares data processing performance of two systems belonging to SQL (PostgreSQL/Postgres XL) and Big Data (Hadoop/Hive) camps on a distributed five-node cluster deployed in cloud. Unlike benchmarks in use (YCSB, TPC), a series of R modules were devised for generating random non-aggregate queries on different subschema (with increasing data size) of TPC-H database. Overall performance of the two systems was compared. Subsequently a number of models were developed for relating performance on the system and also on various query parameters such as the number of attributes in SELECT and WHERE clause, number of joins, number of processing rows etc.
The flood of books, papers, conferences, and workshops about knowledge management, knowledge econ... more The flood of books, papers, conferences, and workshops about knowledge management, knowledge economy, knowledge workers, and knowledge society is incessant. Europe wants to become the best knowledge-based economy in the world. Millions of Euros are pumped every year in national and international research projects dedicated to k-something. This paper is in a certain way against the grain, aiming to demythologize some of the k-excitement.
First Normal Form (1NF) is the only normal form required for a database to be accepted as relatio... more First Normal Form (1NF) is the only normal form required for a database to be accepted as relational. Further normalization-2NF, 3NF, BCNF, 4NF, 5NF, DKNF-is not included in the relational model theory, but it has been considered as a necessary process for diminishing the storage space and for eliminating or reducing the anomalies in insert-update-delete operations. The paradox with 1FN is that it is simultaneously easy to define but not clear to explain in depth. The accepted perspectives on 1FN were seen as largely restrictive, mainly because of atomicity (scalability) of the domains. Recent contributions suggests that 1FN can be seen in a free manner, so that relational model could deal with complex data, especially with objects.
This paper tries to point out some of the promises, quarrels, achievements, and perspectives of t... more This paper tries to point out some of the promises, quarrels, achievements, and perspectives of the forced marriage between the relational and object-oriented data models, from both theoretical and implementation perspectives. As practical O-R achievements, the latest SQL standards (SQL:1999-2008) and Oracle implementation are discussed.
Related to Learning Organization (Chris Argyris, Donald Schon, Peter Senge etc.), but established... more Related to Learning Organization (Chris Argyris, Donald Schon, Peter Senge etc.), but established as a distinct field of study in management/organizational theory through the works of the classics (even though their papers and books are less than ten years old) such as Ikujiro Nonaka, Hitoraka Takeuchi, Thomas Davenport, Laurence Prusak, Verna Allee, Karl Erik Sveiby, Knowledge Management seems to reach its peak of glory, if we consider the number of conferences, books, papers, articles devoted to this topic. For the people who witnessed, ten years ago, the BRP (Business Process Reengineering) fever, there is a strange sensation of deja vu. Is Knowledge Management a fundamental phenomenon in contemporary business environment? Or is it just a passing fad, like so many fads desperately demanded in management theory? No doubt, management theory needs to refresh its gurus' team from time to time. So they rush to exhaustion writing books and dozens of papers each year, trying to ente...
NoSQL data stores, especially the document database systems family (i.e. MongoDB, CouchDB, Couchb... more NoSQL data stores, especially the document database systems family (i.e. MongoDB, CouchDB, Couchbase) implemented a long-awaited feature for many software application developers: a flexible schema for storing and processing data. The idea of schemalessness seemed reasonable and the classic trouble of impedance mismatch finally might have had a solution. Main producers of SQL/relational servers reacted by providing support for storing and querying JSON attributes within tables. This paper approaches the problem of comparing JSON data query performance based on the well-established TPC-H benchmark. Normalized schema was converted into JSON format, and all 22 TPC-H queries were translated to process JSON data. Performance of two popular data servers, Oracle and PostgreSQL, was compared. Resulted testbed can be used to assess the performance of any relational/SQL database engines that supports JSON.
In-memory database systems are among the most recent and most promising Big Data technologies, be... more In-memory database systems are among the most recent and most promising Big Data technologies, being developed and released either as brand new distributed systems or as extensions of old monolith (centralized) database systems. As name suggests, in-memory systems cache all the data into special memory structures. Many are part of the NewSQL strand and target to bridge the gap between OLTP and OLAP into so-called Hybrid Transactional Analytical Systems (HTAP). This paper aims to test the performance of using such type of systems for TPCH analytical workloads. Performance is analyzed in terms of data loading, memory footprint and execution time of the TPCH query set for three in-memory data systems: Oracle, SQL Server and MemSQL. Tests are subsequently deployed on classical on-disk architectures and results compared to in-memory solutions. As in-memory is an enterprise edition feature, associated costs are also considered.
The most prominent Big Data solutions – such as NoSQL systems, Hadoop Frameworks, Spark, etc. – h... more The most prominent Big Data solutions – such as NoSQL systems, Hadoop Frameworks, Spark, etc. – have been open-sourced. Nevertheless, commercial providers have targeted niches of this huge market with products more or less viable and affordable. This paper addresses the problem of benchmarking Big Data platforms with a focus on Oracle Exadata solution provided by one the most important data technologies vendor. Many classical benchmark approaches, such as TPC-H, are based on a predefined set of queries, and consequently they are not prone to predictive modeling. By contrast, for the TPC-H benchmark schema, we generate a set of 500 random queries containing not only tuple filters (WHERE), but also tuple grouping (GROUP BY) and group filters (HAVING), we collected results of the queries execution on four Oracle Exadata settings. Query duration was the outcome variable. Various query parameters, such as the number of joins, the number of attributes of different types within SELECT and ...
Identifying skeletal remains has been and will remain a challenge for forensic doctors and forens... more Identifying skeletal remains has been and will remain a challenge for forensic doctors and forensic anthropologists, especially in disasters with multiple victims or skeletal remains in an advanced stage of decomposition. This study proposes a machine learning method to determine gender starting from morphometric analysis of L1-L5 lumbar vertebrae in a modern Romanian population. The purpose of the present study was to observe whether by using the ML method there is a good predictability of gender in forensic identification based on parameters obtained from the metric analysis of the lumbar spine specific to the Romanian population. This paper offers two models of ML, RF and XGB, each with its own characteristics, and presenting different performance, random forest having the best. For both, we used two metrics (accuracy and roc_auc), the latter being the most used to highlight model performance. The L1-L5 lumbar vertebrae exhibit sexual dimorphism and can be used in gender estimati...
Introduction: Forensic identification practice requires a multitude of methods for positive ident... more Introduction: Forensic identification practice requires a multitude of methods for positive identification, among which the facial reconstruction occupies a high place. For this to be done in the best way possible, it is mandatory that each population have their own database regarding soft tissue thickness. Objective: The present research is focused on creating such a database for the Romanian adult population and on exploring the statistical distribution and correlation of craniometric landmarks, overall, and in groups defined according to sex and weight categories. This might provide support in estimating the values of missing landmarks or other variables (e.g., age, sex) for unidentified bodies. Methods: Twelve craniometric landmarks were measured on 100 cadavers less than 24 h after death. Results and discussions: Of the 12 anatomical landmarks, only one appears to manifest statistically significant differences between sexes, and seven landmarks appear to vary significantly amon...
Energy management systems from Romania do not have the capabilities of energy specific management... more Energy management systems from Romania do not have the capabilities of energy specific management due to lack of technology for real-time monitoring. As was the case in many other countries, the advent of smart metering technology will increase the level of energy data significantly. Therefore, the purpose of this paper is to present solutions that need to be taken to solve problems linked with the increasing amount of data recorded by sensors. For a better demonstration of theoretical elements exposed, we considered a data warehouse specific to utility companies. Section 2 of this article defines the three widely used parallel data warehouse architectures, while in Section 3 we clarify what architecture is suited to develop a data warehouse in the cloud. In the last part we transposed our tables in a "shared-nothing" architecture, trying to analyze queries performance.
Proceedings of the 19th International Conference on Computer Systems and Technologies, 2018
Big Data technologies brought the idea of parallel processing on cheaper commodity servers. When ... more Big Data technologies brought the idea of parallel processing on cheaper commodity servers. When dealing with huge amount of data, instead of migrating to more performant and costly hardware platforms, or buying resources in cloud, it is more affordable to add a number of cheaper servers as nodes for data processing and/or storage. NoSQL data stores, Hadoop ecosystems, NewSQL platforms have proved viable for Big Data storage and processing. In this paper we were concerned with setting up a platform for big data processing using commodity workstations. Many small and medium sized companies have limited resources and their workstations remain unused for more than 12 hours a day. Here Beowulf Cluster Computing could prove useful. Apache Impala was installed as part of a Hadoop distribution on a 9-node cluster. Three TPC-H database schema were loaded for the scale factors of 1, 2 and 10GB. A series of 100 SQL queries were randomly generated and executed for each scale factor. Results we...
Big Data systems manage and process huge volumes of data constantly generated by various technolo... more Big Data systems manage and process huge volumes of data constantly generated by various technologies in a myriad of formats. Big Data advocates (and preachers) have claimed that, relative to classical, relational/SQL Data Base Management Systems, Big Data technologies such as NoSQL, Hadoop and in-memory data stores perform better. This paper compares data processing performance of two systems belonging to SQL (PostgreSQL/Postgres XL) and Big Data (Hadoop/Hive) camps on a distributed five-node cluster deployed in cloud. Unlike benchmarks in use (YCSB, TPC), a series of R modules were devised for generating random non-aggregate queries on different subschema (with increasing data size) of TPC-H database. Overall performance of the two systems was compared. Subsequently a number of models were developed for relating performance on the system and also on various query parameters such as the number of attributes in SELECT and WHERE clause, number of joins, number of processing rows etc.
The flood of books, papers, conferences, and workshops about knowledge management, knowledge econ... more The flood of books, papers, conferences, and workshops about knowledge management, knowledge economy, knowledge workers, and knowledge society is incessant. Europe wants to become the best knowledge-based economy in the world. Millions of Euros are pumped every year in national and international research projects dedicated to k-something. This paper is in a certain way against the grain, aiming to demythologize some of the k-excitement.
First Normal Form (1NF) is the only normal form required for a database to be accepted as relatio... more First Normal Form (1NF) is the only normal form required for a database to be accepted as relational. Further normalization-2NF, 3NF, BCNF, 4NF, 5NF, DKNF-is not included in the relational model theory, but it has been considered as a necessary process for diminishing the storage space and for eliminating or reducing the anomalies in insert-update-delete operations. The paradox with 1FN is that it is simultaneously easy to define but not clear to explain in depth. The accepted perspectives on 1FN were seen as largely restrictive, mainly because of atomicity (scalability) of the domains. Recent contributions suggests that 1FN can be seen in a free manner, so that relational model could deal with complex data, especially with objects.
This paper tries to point out some of the promises, quarrels, achievements, and perspectives of t... more This paper tries to point out some of the promises, quarrels, achievements, and perspectives of the forced marriage between the relational and object-oriented data models, from both theoretical and implementation perspectives. As practical O-R achievements, the latest SQL standards (SQL:1999-2008) and Oracle implementation are discussed.
Related to Learning Organization (Chris Argyris, Donald Schon, Peter Senge etc.), but established... more Related to Learning Organization (Chris Argyris, Donald Schon, Peter Senge etc.), but established as a distinct field of study in management/organizational theory through the works of the classics (even though their papers and books are less than ten years old) such as Ikujiro Nonaka, Hitoraka Takeuchi, Thomas Davenport, Laurence Prusak, Verna Allee, Karl Erik Sveiby, Knowledge Management seems to reach its peak of glory, if we consider the number of conferences, books, papers, articles devoted to this topic. For the people who witnessed, ten years ago, the BRP (Business Process Reengineering) fever, there is a strange sensation of deja vu. Is Knowledge Management a fundamental phenomenon in contemporary business environment? Or is it just a passing fad, like so many fads desperately demanded in management theory? No doubt, management theory needs to refresh its gurus' team from time to time. So they rush to exhaustion writing books and dozens of papers each year, trying to ente...
NoSQL data stores, especially the document database systems family (i.e. MongoDB, CouchDB, Couchb... more NoSQL data stores, especially the document database systems family (i.e. MongoDB, CouchDB, Couchbase) implemented a long-awaited feature for many software application developers: a flexible schema for storing and processing data. The idea of schemalessness seemed reasonable and the classic trouble of impedance mismatch finally might have had a solution. Main producers of SQL/relational servers reacted by providing support for storing and querying JSON attributes within tables. This paper approaches the problem of comparing JSON data query performance based on the well-established TPC-H benchmark. Normalized schema was converted into JSON format, and all 22 TPC-H queries were translated to process JSON data. Performance of two popular data servers, Oracle and PostgreSQL, was compared. Resulted testbed can be used to assess the performance of any relational/SQL database engines that supports JSON.
In-memory database systems are among the most recent and most promising Big Data technologies, be... more In-memory database systems are among the most recent and most promising Big Data technologies, being developed and released either as brand new distributed systems or as extensions of old monolith (centralized) database systems. As name suggests, in-memory systems cache all the data into special memory structures. Many are part of the NewSQL strand and target to bridge the gap between OLTP and OLAP into so-called Hybrid Transactional Analytical Systems (HTAP). This paper aims to test the performance of using such type of systems for TPCH analytical workloads. Performance is analyzed in terms of data loading, memory footprint and execution time of the TPCH query set for three in-memory data systems: Oracle, SQL Server and MemSQL. Tests are subsequently deployed on classical on-disk architectures and results compared to in-memory solutions. As in-memory is an enterprise edition feature, associated costs are also considered.
The most prominent Big Data solutions – such as NoSQL systems, Hadoop Frameworks, Spark, etc. – h... more The most prominent Big Data solutions – such as NoSQL systems, Hadoop Frameworks, Spark, etc. – have been open-sourced. Nevertheless, commercial providers have targeted niches of this huge market with products more or less viable and affordable. This paper addresses the problem of benchmarking Big Data platforms with a focus on Oracle Exadata solution provided by one the most important data technologies vendor. Many classical benchmark approaches, such as TPC-H, are based on a predefined set of queries, and consequently they are not prone to predictive modeling. By contrast, for the TPC-H benchmark schema, we generate a set of 500 random queries containing not only tuple filters (WHERE), but also tuple grouping (GROUP BY) and group filters (HAVING), we collected results of the queries execution on four Oracle Exadata settings. Query duration was the outcome variable. Various query parameters, such as the number of joins, the number of attributes of different types within SELECT and ...
Uploads
Papers by Marin Fotache