30988 research outputs found
Sort by
Accent Imitation as A Voice Disguise Method: A Study on the Articulators' Movements Data
No full textInhimillinen herkkyys : kadotettu voimavara tehokkuuden aikakaudella? Moninaisen toimijuuden merkitys matkalla kohti kestävämpää hyvinvointia
No full textPoor Leadership Through the Eyes of Frontline Managers. Experiences from the Social and Health Care Sector
No full textIntertekstuaaliset viittaukset ja reaaliat neljän Vladimir Vysotskin runon suomennoksissa
No full textCHILDREN ABROAD: A RELATIONAL ANALYSIS OF FINNISH CHILD PROTECTION AND WELFARE IN TRANSNATIONAL CONTEXTS [CARELA] Case files
No full textThe legal framework created to harmonise the rules governing cross-border child protection rests mainly on three instruments. Within the EU, child protection is primarily ensured by the Brussels IIter Regulation, which replaced the Brussels IIa Regulation as of 1 August 2022. In addition to the Regulation, all EU Member States have ratified the Hague Convention on Child Abduction 1980, and the Hague Convention on the Protection of Children 1996. All of these instruments base the jurisdiction to issue protective measures on the habitual residence of the child. They also establish a system of central authorities (CAs), who operate on the national level and are responsible for sharing good practice and information on national laws and practices and cooperating in cross-border cases. In Finland, the CA is the Ministry of Justice.
The contact point function of the CAs means that research into the actual cases of the CA can offer invaluable information as to how cross-border child protection operates in practice. It can also indicate factors enhancing or precluding effective protection of children in transnational situations. This dataset is based on detailed notes made of case files from the Finnish Ministry of Justice. It covers cases of child protection in 2012–2020 (n=193) and child abduction in 2016–2018 (n=53). Of the child protection cases, the requesting authority was Finnish in 98 cases and in 95 cases the request came from a foreign CA
"I have never been 100 % sure if I'm in the right profession": Teacher students' and novice teachers' thoughts about a career change
No full textServant Leadership as Support for the Adoption of Artificial Intelligence in Expert Organizations
No full text
An improved brain-motivated network for forecasting day-ahead stock prices of electricity companies
No full textGiven the complexity of the trading market, stock prices of electricity companies present nonlinear, non-stationary, and random fluctuations, resulting in its high-precision forecasting being a challenging task. Deep learning, particularly in bionically-inspired networks, has shown great potential in time series forecasting. To this end, this paper proposes an improved brain-motivated network with GELU function and residual connection for forecasting the day-ahead stock prices of electricity companies. Specifically, from a macro view, the improved brain-motivate network effectively simulates gate, parallel handling, and collaboration functions in the brain, inheriting some of the superior analytical capabilities of the biological brain. From a micro view, the GELU function is first adopted to extract nonlinear features, which not only avoids the gradient vanishing of Tanh function but also deals with the learning issue of ReLU function when the input is negative. Moreover, the residual connection is utilized to convey information between the shallow and deep layers, making full use of information for mining deeply hidden features. In summary, the cooperation of the above components improves the forecasting accuracy. Experimental results and analysis of two case study from Finnish electricity companies show that the improved brain-motivated network outperforms 18 baselines with improved average mean absolute percentage errors of 13.28 % and 8.56 %, thus helping stakeholders make informed decisions and gain profitable returns