Content-Type: text/html;
charset=utf-8
D=C3=A9but du message r=C3=A9exp=C3=A9di=C3=A9 :
De: =
Eduardo Ferm=C3=A9 <eduardo.ferme@staff.uma.pt>
Eduardo Ferm=C3=A9 <eduardo.ferme@staff.uma.pt>
Objet: =
1 Posdoc Position =
– Applied machine Learning – formal methods applied to artificial =
intelligence
1 Posdoc Position =
– Applied machine Learning – formal methods applied to artificial =
intelligence
Date: 21=
novembre 2018 =C3=A0 15:38:05 UTC+1
novembre 2018 =C3=A0 15:38:05 UTC+1
=C3=80: =
Eduardo Ferm=C3=A9 <ferme@uma.pt>
Eduardo Ferm=C3=A9 <ferme@uma.pt>
———-
1 Posdoc Position – Applied machine =
Learning – formal methods applied to artificial intelligence
1 Posdoc Position – Applied machine =
Learning – formal methods applied to artificial intelligence
The =
NOVA LINCS Knowledge-Based Systems Group at Madeira University is =
seeking candidates for a PostDoc position to work within the scope of =
the following project:
NOVA LINCS Knowledge-Based Systems Group at Madeira University is =
seeking candidates for a PostDoc position to work within the scope of =
the following project:
Project BRANT – Belief =
Revision applied to Neurorehabilitation Therapy =
(PTDC/CCI-COM/30990/2017)
Revision applied to Neurorehabilitation Therapy =
(PTDC/CCI-COM/30990/2017)
Cognitive=
deficits are common after brain injury, dementia and in normal =
cognitive
deficits are common after brain injury, dementia and in normal =
cognitive
decline due to aging. Current =
cognitive rehabilitation therapy has been shown to be the
cognitive rehabilitation therapy has been shown to be the
most effective way to address this problem. =
However, a) they are not adaptive for every
However, a) they are not adaptive for every
patient and b) have a high cost, and is =
usually implemented in clinical environments.
usually implemented in clinical environments.
The Task Generator (TG) is a free tool for the =
generation of cognitive training tasks.
generation of cognitive training tasks.
However, TG is not designed to adapt and =
monitor the evolution of the patient. Here we
monitor the evolution of the patient. Here we
propose BRaNT, an enhancement of TG with =
Artificial Intelligence modules, gamification
Artificial Intelligence modules, gamification
and remote monitoring capabilities to enable =
Health Professionals to provide long-term
Health Professionals to provide long-term
personalized cognitive rehabilitation therapy =
at home. BRaNT is an interdisciplinary
at home. BRaNT is an interdisciplinary
effort that addresses scientific limitations =
of current practices as well as provides
of current practices as well as provides
solutions towards the sustainability of health =
systems and contributes towards the
systems and contributes towards the
improvement of quality of life of patients.
Candidates should hold a PhD. or equivalent degree. =
Suitable disciplines include Computer Science, Artificial Intelligence, =
Mathematics, Machine Learning and other related areas. A strong =
publication record is considered a plus. Candidates are expected to be =
fluent in spoken and written English. PostDocs can, optionally, =
supervise MSc and/or PhD projects.
Suitable disciplines include Computer Science, Artificial Intelligence, =
Mathematics, Machine Learning and other related areas. A strong =
publication record is considered a plus. Candidates are expected to be =
fluent in spoken and written English. PostDocs can, optionally, =
supervise MSc and/or PhD projects.
It is desirable that the =
candidate has the following experience/skills:
candidate has the following experience/skills:
1. =
Knowledge in logic or formal methods applied to artificial =
intelligence.
Knowledge in logic or formal methods applied to artificial =
intelligence.
2. Knowledge in Applied =
Machine Learning and Data Science.
Machine Learning and Data Science.
3. =
Strong programming skills
Strong programming skills
Task will include, among =
others, to participate in a interdisciplinary team to perform:
others, to participate in a interdisciplinary team to perform:
(a) =
Patient Profile Generator
Patient Profile Generator
Based =
in normalized data in a database, we will create a assistive tool to =
create a patient profile which represents his/her cognitive competences. =
This training profile will be created by operationalizing tasks =
according to how their different parameters impact different cognitive =
domains (Attention, Memory, Executive Functions, Language, Processing =
Speed). This will be achieved by means of a participatory design =
methodology involving 20 rehabilitation experts and end-user groups. =
Through modeling we expect to quantitatively determine how independent =
variables (task parameters) impact each of the profile domains. For each =
patient, this profiling and parameter matching process will result in =
the selection of a number of tasks that are adjusted to the patient and =
delivered as the prescribed training.
in normalized data in a database, we will create a assistive tool to =
create a patient profile which represents his/her cognitive competences. =
This training profile will be created by operationalizing tasks =
according to how their different parameters impact different cognitive =
domains (Attention, Memory, Executive Functions, Language, Processing =
Speed). This will be achieved by means of a participatory design =
methodology involving 20 rehabilitation experts and end-user groups. =
Through modeling we expect to quantitatively determine how independent =
variables (task parameters) impact each of the profile domains. For each =
patient, this profiling and parameter matching process will result in =
the selection of a number of tasks that are adjusted to the patient and =
delivered as the prescribed training.
(b) =
Belief Revision adaptation engine (BR-E: Adaptation).
Belief Revision adaptation engine (BR-E: Adaptation).
The aim of this WP is to create the =
computational infrastructure based on the Belief Revision framework that =
will
computational infrastructure based on the Belief Revision framework that =
will
enable the accurate =
generation of cognitive profiles for patients, as well as the update of =
these profiles over time to accurately capture the patients’ evolution =
over time and the consequent adaptation of cognitive training (Belief =
Revision adaptation engine).
generation of cognitive profiles for patients, as well as the update of =
these profiles over time to accurately capture the patients’ evolution =
over time and the consequent adaptation of cognitive training (Belief =
Revision adaptation engine).
The BR-E must be capable to detects =
inconsistencies (i.e., formal contradictions) and incoherencies (i.e., =
results between different assessment that are outside of expected =
differences). In case of normal evolution/involution the system will be =
update the profle accordingly and cycle the process with the updated =
profile.
inconsistencies (i.e., formal contradictions) and incoherencies (i.e., =
results between different assessment that are outside of expected =
differences). In case of normal evolution/involution the system will be =
update the profle accordingly and cycle the process with the updated =
profile.
Gross Salary: Approx 36K =E2=82=AC x year.
The =
positions are available from 1/4/2018. The positions are fixed-term and =
filled initially for 1 year with the option to be renewed until the end =
of the project (30/9/21).
positions are available from 1/4/2018. The positions are fixed-term and =
filled initially for 1 year with the option to be renewed until the end =
of the project (30/9/21).
Applicants should express =
their interest by sending the following documents by email to Eduardo =
Ferm=C3=A9 (ferme@uma.pt) no later =
than 20/12/2018, according to the instructions below:
their interest by sending the following documents by email to Eduardo =
Ferm=C3=A9 (ferme@uma.pt) no later =
than 20/12/2018, according to the instructions below:
– a cover letter briefly describing your =
background and interests;
background and interests;
– a =
Curriculum Vitae (with contact details);
Curriculum Vitae (with contact details);
– names and contact information (email) of at =
least two references;
least two references;
– a =
link to electronic copy of Ph.D. Thesis (if available);
link to electronic copy of Ph.D. Thesis (if available);
– links to electronic copies of 1-2 most =
relevant publications.
relevant publications.
—
Eduardo Ferm=C3=A9
Faculty of Exact =
Sciences and Engineering
Sciences and Engineering
University of Madeira
Campus Universit=C3=A1rio da =
Penteada 9000-390 Funchal, Madeira, Portugal.
Penteada 9000-390 Funchal, Madeira, Portugal.
tel. + 351 291 705179
e-mail: ferme@uma.pt http://www.cee.uma.pt/ferme
____as____