60 research outputs found
Vuorovaikutteisia koneoppimismenetelmiä asiantuntijatiedon tuomiseen laskennallisiin malleihin täsmälääketieteen päätöksenteon tueksi
No full textThis thesis develops human-in-the-loop machine learning methods that aim at improving the performance of a machine learning model in precision medicine tasks. Many problems in precision medicine are still difficult for machine learning due to lack of data, and human experts' knowledge can provide a valuable source of information to reduce a model's prediction error and uncertainty. Such expert knowledge elicitation requires methods that address the following problems: How to leverage indirect expert knowledge instead of querying labels as in active learning, how to make the interaction less laborious to the expert than in traditional prior elicitation, and how to select the interaction so that it is the most beneficial to the prospective task of the model.
The first contribution of the thesis is to develop an interactive knowledge elicitation method for "small n large p" problems where data is insufficient, that allows even a small amount of sequentially chosen noisy, indirect feedback from an expert to complement the data and improve the accuracy of the model's predictions. The effectiveness of the method is evaluated in a user-study. The method is further extended to a high-dimensional genomics prediction task where we demonstrate, for the first time, how sequentially selected domain expert's feedback improves personalized prediction of the cancer cell's sensitivity to drugs.
The second main contribution of the thesis is to introduce two goal-oriented data acquisition strategies that aim at selecting queries that are maximally useful for a prospective task where the model is to be used: First, targeted Bayesian optimal experimental design to increase the accuracy of a single personalized prediction, and second, active learning that takes the down-the-line decision-making task into account by modeling the probability of a wrong decision.
The last part of this thesis applies human-in-the-loop methods to a new, promising and yet unexplored application domain of de novo molecular design. The last contribution is how the goal of molecule generation can be inferred via human-in-the-loop interaction, to make an adaptive objective function to a reinforcement learning algorithm, so that the resulting system generates more molecules that match the user's goal.Tässä väitöskirjassa kehitetään vuorovaikutteisia koneoppimismenetelmiä, joilla pyritään parantamaan koneoppimismallin suorituskykyä täsmälääketieteen tehtävissä. Monet täsmälääketieteen ongelmat ovat edelleen vaikeita koneoppimiselle datan puutteen vuoksi, mutta asiantuntijoiden näkemys tarjoaa arvokkaan tietolähteen, jolla voidaan vähentää mallien ennustevirhettä ja epävarmuutta. Tällaisen asiantuntijatiedon hankkiminen edellyttää menetelmiä, jotka ratkaisevat seuraavat ongelmat: Miten hyödyntää myös epäsuoraa asiantuntijatietoa sen sijaan, että ihminen paljastaa mallille oikeat vastaukset kuten aktiivioppimisessa, miten tehdä vuorovaikutuksesta asiantuntijalle vähemmän työlästä kuin perinteinen priorijakaumien määrittely, ja miten valita asiantuntijalle esitettävät kysymykset siten, että niistä on eniten hyötyä mallin tulevan tehtävän kannalta.
Väitöskirjassa kehitetään interaktiivinen menetelmä asiantuntijatiedon tuomiseen malliin, jolla voidaan ratkaista ns. "pieni n suuri p" -ongelmia, joissa dataa ei ole riittävästi. Menetelmän avulla pienikin määrä aktiivisesti valittua asiantuntijan antamaa palautetta parantaa mallin ennusteiden tarkkuutta kun se yhdistetään data-analyysiin. Menetelmän tehokkuus osoitetaan käyttäjätutkimuksessa. Tämä menetelmä laajennetaan myös korkeaulotteiseen genomiikan ennustustehtävään ja näytetään ensimmäistä kertaa, että asiantuntijan palaute parantaa yksilöllisiä ennusteita syöpäsolujen lääkeherkkyydestä.
Väitöskirjan toinen keskeinen tulos on kahden tavoitteellisen aktiivioppimisstrategian kehittäminen, joilla pyritään valitsemaan asiantuntijalle esitettävät kysymykset niin, että ne ovat mahdollisimman hyödyllisiä tehtävässä, jossa koneoppimismallia on tarkoitus käyttää. Ensimmäinen menetelmä on kohdennettu Bayesilainen optimaalinen koesuunnittelu, jolla pyritään lisäämään yksittäisen yksilöllistetyn ennusteen tarkkuutta. Toinen on päätöksentekoon keskittyvä aktiivioppimismenetelmä, joka ottaa huomioon edessä olevan päätöksentekotehtävän mallintamalla väärän päätöksen todennäköisyyden ja minimoimalla sitä.
Väitöskirjan viimeisessä osassa sovelletaan vuorovaikutteisia koneoppimismenetelmiä uuteen, lupaavaan ja vielä tutkimattomaan sovellusalueeseen, de novo -molekyylisuunnitteluun. Tuloksena on menetelmä, jossa algoritmi vuorovaikuttaa kemistin kanssa ja päättelee tämän palautteen perusteella molekyylisuunnittelutehtävän tavoitteen. Tällöin vahvistusoppimisalgoritmille voidaan luoda adaptiivinen tavoitefunktio niin, että järjestelmä tuottaa enemmän kemistin tavoitetta vastaavia molekyylejä
Organisational Entrepreneurs in the Public Sectors - Social Capital and Gender
No full text This book explores social capital as the multiple relationships between gender, management and entrepreneurship. Human resources are the social capital of a firm and business life, based on trust as well as on expertise, values and cultural diversity. This calls for cross-cultural knowledge and an understanding of gender issues and individual differences in the social capital of the firm and society. The dialogue between women entrepreneurship and social capital theory and research has its special place among other women entrepreneurship books, the number of which has lately increased. It strengthens still in some respect the fragmented voice of women entrepreneurship research by providing a landscape of women entrepreneurs as creators of, and created by, social capital. It indicates how women entrepreneurs appear to have a special position in forming, developing and reorganizing the social capital in the business world. In its eleven chapters, twenty-six researchers representing a variety of disciplines from different parts of the world are presenting findings on diverse aspects of the dialogue between women entrepreneurship and social capital.As a consequence the central concepts, social capital, entrepreneurship and gender, are given a variety of meanings. Women entrepreneurs and business owners regardless of their cultural context, branch and education provide interesting ideas to the global debate on equality and social capital. Iiris Aaltio (Ph.D., Econ.) is Professor of Management at the School of Economics and Business in the University of Jyvaskyla, Finland. She has worked as a Visiting Scholar at the University of Massachusetts, MIT, the University of St. Mary's and Auckland University. Her research is focused on organization culture, gender and organizational entrepreneurship. Paula Kyro (Ph.D. Econ and Ph.D. Educ.) has worked as Professor in both Education and Entrepreneurship. She is currently working as Professor of Entrepreneurship Education at the Helsinki School of Economics, Finland. She has also worked as a Visiting Professor in entrepreneurship at Jonkoping International Business School. Besides women entrepreneurship, her research interests are in the methodology and entrepreneurship education, dynamics of entrepreneurial and enterprising learning and readiness as well as in virtual learning.Elisabeth Sundin (Ph.D, Business Administration and Management) is Professor in Business Administration and Management at Linkoping University, Sweden. She has worked as a Professor at the National Institute of Working Life and also at Jonkoping International Business School. She has done research on SMEs and gender and is now combining the two in research on the reorganisation of the public sector.</p
Female organisational entrepreneurs in the care sector : the importance of social capital and gender
No full textThis book explores social capital as the multiple relationships between gender, management and entrepreneurship. Human resources are the social capital of a firm and business life, based on trust as well as on expertise, values and cultural diversity. This calls for cross-cultural knowledge and an understanding of gender issues and individual differences in the social capital of the firm and society. The dialogue between women entrepreneurship and social capital theory and research has its special place among other women entrepreneurship books, the number of which has lately increased. It strengthens still in some respect the fragmented voice of women entrepreneurship research by providing a landscape of women entrepreneurs as creators of, and created by, social capital. It indicates how women entrepreneurs appear to have a special position in forming, developing and reorganizing the social capital in the business world. In its eleven chapters, twenty-six researchers representing a variety of disciplines from different parts of the world are presenting findings on diverse aspects of the dialogue between women entrepreneurship and social capital.As a consequence the central concepts, social capital, entrepreneurship and gender, are given a variety of meanings. Women entrepreneurs and business owners regardless of their cultural context, branch and education provide interesting ideas to the global debate on equality and social capital. Iiris Aaltio (Ph.D., Econ.) is Professor of Management at the School of Economics and Business in the University of Jyvaskyla, Finland. She has worked as a Visiting Scholar at the University of Massachusetts, MIT, the University of St. Mary's and Auckland University. Her research is focused on organization culture, gender and organizational entrepreneurship. Paula Kyro (Ph.D. Econ and Ph.D. Educ.) has worked as Professor in both Education and Entrepreneurship. She is currently working as Professor of Entrepreneurship Education at the Helsinki School of Economics, Finland. She has also worked as a Visiting Professor in entrepreneurship at Jonkoping International Business School. Besides women entrepreneurship, her research interests are in the methodology and entrepreneurship education, dynamics of entrepreneurial and enterprising learning and readiness as well as in virtual learning.Elisabeth Sundin (Ph.D, Business Administration and Management) is Professor in Business Administration and Management at Linkoping University, Sweden. She has worked as a Professor at the National Institute of Working Life and also at Jonkoping International Business School. She has done research on SMEs and gender and is now combining the two in research on the reorganisation of the public sector.</p
AI-assisted curriculum learning
No full textDeep reinforcement learning is widely applied in de novo molecular design to generate molecules with desired properties. This technique often has a sparse reward problem since the target properties usually exist for the minority of the generated molecules. With a sparse reward, the agent in a de novo design tool may fail to begin learning and waste much time exploring areas in the vast chemical space that are far away from the target area. A recent study successfully applied curriculum learning to mitigate the sparse reward problem. However, a chemist must hand-craft a curriculum for the generative agents, which requires domain knowledge and is time-consuming, especially as tasks grow in complexity. This thesis applies an AI assistance framework to assist in a curriculum design task by recommending actions. The AI assistant infers the private information of the chemist, including design objective function and chemist’s biases. Then, the AI tries to convince the chemist to adopt its advice. The chemist is free to choose action after receiving advice. This setting presents a significant improvement in AI safety. We demonstrate this method with a simulated chemist in a de novo design task, where the generated molecules should be predicted to be active against the dopamine type 2 receptor (DRD2). Our experiments show that the AI-assisted curriculum learning achieves a pronounced improvement on the sparse property (DRD2) and significantly outperforms unassisted curriculum learning
Women, Entrepreneurship and Social Capital : A Dialogue and Construction
No full textThis book explores social capital as the multiple relationships between gender, management and entrepreneurship. Human resources are the social capital of a firm and business life, based on trust as well as on expertise, values and cultural diversity. This calls for cross-cultural knowledge and an understanding of gender issues and individual differences in the social capital of the firm and society. The dialogue between women entrepreneurship and social capital theory and research has its special place among other women entrepreneurship books, the number of which has lately increased. It strengthens still in some respect the fragmented voice of women entrepreneurship research by providing a landscape of women entrepreneurs as creators of, and created by, social capital. It indicates how women entrepreneurs appear to have a special position in forming, developing and reorganizing the social capital in the business world. In its eleven chapters, twenty-six researchers representing a variety of disciplines from different parts of the world are presenting findings on diverse aspects of the dialogue between women entrepreneurship and social capital. As a consequence the central concepts, social capital, entrepreneurship and gender, are given a variety of meanings. Women entrepreneurs and business owners regardless of their cultural context, branch and education provide interesting ideas to the global debate on equality and social capital.</p
- …
